KR20200020915A - Image data encoding method and apparatus and Image data decoding method and apparatus - Google Patents

Abstract

Embodiments of the present application disclose an image data encoding method and apparatus, and an image data decoding method and apparatus, which relate to the field of image processing and solve the problem of relatively high coding complexity. The decoding method comprises: obtaining a bitstream comprising picture data; Obtaining the node split mode information of the first-level coding tree by parsing the bitstream and obtaining the node split mode information of the second-level coding tree by parsing the bitstream-the second level coding The node splitting mode information of the tree indicates a splitting mode corresponding to a first node of the second-level coding tree, and the splitting mode corresponding to the first node is a candidate-split-mode corresponding to the first node. One mode in the set, wherein the candidate-division-mode set corresponding to the first node is determined according to a first preset dividing condition, wherein the first preset dividing condition is determined by the first node in a target dividing mode. Used to indicate whether to restrict splitting; If the splitting mode corresponding to the first node is no further splitting, parsing the bitstream to obtain encoding information of the first node; And decoding and reconstructing the coding unit based on the encoding information of the first node to obtain a picture corresponding to the picture data.

Description

Image data encoding method and apparatus and Image data decoding method and apparatus

Embodiments of the present application relate to the field of image processing, and in particular, to a method and apparatus for encoding image data and a method and apparatus for decoding image data.

In video codec standard H.265, picture frames are divided into coding tree units (CTUs) that do not overlap each other, and each CTU is used as a root node of a quad-tree (QT). do. Each CTU is recursively divided into multiple leaf nodes according to the QT structure. Each node of the QT structure corresponds to one picture region. When a node is no longer split, the node is called a leaf node, and the picture region corresponding to that node forms a coding unit (CU). Therefore, H.265 can be considered as a process of dividing the CTU into one CU group. The splitting mode for dividing the CTU into CU groups corresponds to one coding tree.

Joint Exploration Model (JEM) by Joint Exploration team on Future Video Coding (JVET) proposes QTBT splitting mode. Specifically, the nodes of the first-level coding tree use the QT splitting mode, and the nodes of the second-level coding tree use the binary tree (BT) splitting mode. Binary binary split "and" vertical binary split ". Specifically, the CTU is partitioned first in the QT partitioning mode to obtain some QT leaf nodes; The QT leaf node may be split in the BT split mode. In QTBT splitting mode, the CU shape is more versatile to better adapt to the content of the local picture.

However, for each node, the encoder device generally calculates the rate distortion cost (RD cost) of each split mode that can be used by the node, compares the calculated RD cost, and compares the minimum RD cost. The split mode corresponding to must be determined as the split mode of the node. Thus, for each node, the encoder device needs to calculate the RD costs of the plurality of split modes, and consequently the encoding complexity is relatively high.

Embodiments of the present application provide an image data encoding method and apparatus and an image data decoding method and apparatus to solve the problem of relatively high coding complexity.

In order to achieve the above object, the following technical solutions are used in the embodiments of the present application.

According to a first aspect, a method of decoding image data is provided. After acquiring the bitstream including the image data, the decoder device parses the bitstream to obtain node dividing mode information of the first-level coding tree and node dividing mode information of the second-level coding tree. The level coding tree indicates a split mode corresponding to the first node of the second-level coding tree, the split mode corresponding to the first node is one mode in the candidate-split-mode set corresponding to the first node, and the first The candidate-mode set corresponding to the node is determined according to the first preset splitting condition, the first preset splitting condition is used to indicate whether the first node restricts splitting in the target splitting mode, and the target splitting mode is horizontal. At least one of binary division, horizontal ternary division, vertical binary division, and vertical ternary division. If the splitting mode corresponding to the first node is non-additional splitting, the decoder device parses the bitstream to obtain encoding information of the first node, where the first node corresponds to one coding unit (CU). In this way, the decoder device can decode and reconstruct the coding unit based on the encoding information of the first node to obtain a picture corresponding to the picture data. In the present application, the root node of the first-level coding tree corresponds to one CTU, and the root node of the first-level coding tree is a node split of the root node of the first-level coding tree and the first-level coding tree. It is identified using the node splitting mode corresponding to the information. The root node of the second-level coding tree is a leaf node of the first-level coding tree.

In this embodiment of the present application, the splitting mode corresponding to the first node of the second-level coding tree is one of splitting modes determined according to the first preset splitting condition, and the first preset splitting condition is determined by the first node. Split in the target division mode, the target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. The first preset partitioning condition reduces the quantity of partitioning modes corresponding to the first node, effectively reducing the decoding complexity.

Optionally, in a possible implementation of the present application, the node splitting mode corresponding to the first-level coding tree is different from the node splitting mode corresponding to the second-level coding tree. The node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary tree splitting.

Optionally, in another possible implementation of the present application, the first preset splitting subcondition, the first preset splitting subcondition, the second preset splitting subcondition, the third preset splitting subcondition, the fourth preset splitting subcondition, And at least one of a fifth preset division sub-condition, a sixth preset division sub-condition, a seventh preset division sub-condition, an eighth preset division sub-condition, and a ninth preset division sub-condition. A first preset division sub-condition is as follows: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to a first preset threshold, The corresponding candidate-division-mode set does not include the horizontal binary division or the horizontal ternary division. The width to height ratio of the image area corresponding to the first node is a ratio of the width of the image area corresponding to the first node to the height of the image area corresponding to the first node. A second preset division subcondition is as follows: if the height-to-width ratio of the picture area corresponding to the first node is greater than or equal to a second preset threshold, the candidate-division-mode corresponding to the first node. The set does not include the vertical binary division or the vertical ternary division. The height to width ratio of the image area corresponding to the first node is a ratio of the height of the image area corresponding to the first node to the width of the image area corresponding to the first node. A third preset division subcondition is as follows: an area of the picture area corresponding to the first node with respect to an area of the picture area corresponding to the leaf node of the first-level coding tree to which the first node belongs. If the ratio of the ratio is less than or equal to a third preset threshold, the candidate-division-mode set corresponding to the first node is the horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. Does not include A fourth preset partitioning subcondition is as follows: if the partitioning mode corresponding to the first node includes a first partitioning mode, decoding of the first child node of the first node is performed by the second node of the first node. Later than decoding of the child node, the splitting mode corresponding to the second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode. The first division mode is the horizontal binary division or the vertical binary division. The fifth preset division sub-condition is as follows: the division mode corresponding to the first node includes a second division mode, and the immunity of the image region corresponding to the first child node of the first node is increased. If the smallest among the immunities of the picture regions corresponding to the three child nodes of the first node, the candidate-division-mode set corresponding to the first child node does not include the second division mode. The second division mode is the horizontal ternary division or the vertical ternary division. A sixth preset partitioning subcondition is as follows: the partitioning mode corresponding to the first node includes the second partitioning mode, and the immunity of the picture region corresponding to the first child node of the first node. If the immunity of the picture region corresponding to the three child nodes of the first node is the largest, the candidate-division-mode set corresponding to the first child node does not include the second division mode. A seventh preset division sub-condition is as follows: if the ratio of the width of the picture area corresponding to the first node to a preset side length of a minimum CU is less than or equal to a third preset threshold, the The candidate-division-mode set corresponding to the first node does not include the vertical ternary division, or the ratio of the height of the picture area corresponding to the first node to the preset side length of the minimum CU is If less than or equal to a third preset threshold, the candidate-division-mode set corresponding to the first node includes the horizontal ternary division. An eighth preset partitioning subcondition is as follows: if the immunity of the picture region corresponding to the first node is less than or equal to a fourth preset threshold, the candidate-division-mode set corresponding to the first node is The horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. A ninth preset divisional subcondition is as follows: the division mode corresponding to the first node is the second division mode, and the immunity of the image region corresponding to the first child node of the first node is If the immunity of the picture region corresponding to the three child nodes of the first node is the largest, the candidate-division-mode set corresponding to the first child node does not include the first partitioning mode. The division direction of the first division mode is the same as the division direction of the second division.

Optionally, in another possible implementation of the present application, node splitting mode information of the second-level coding tree includes first information, second information, and third information. The first information is used to indicate whether to further split the first node, the second information is used to indicate the direction in which the first node is split, and the third information is used to indicate the mode in which the first node is split. In this scenario, the method of "parsing the bitstream to obtain node splitting mode information of the second-level coding tree" is as follows: The bitstream is a candidate corresponding to the first node according to the first preset splitting condition. Parsed to determine a split-mode set; The bitstream is parsed based on the candidate-division-mode set corresponding to the first node to determine the first information, the second information, and the third information.

According to a parsing method of obtaining first information, second information and third information by a decoder device based on a candidate-division-mode set corresponding to a first node, the parsing method includes: a value of the first information, a value of the second information, The value of the third information may be parsed from the bitstream or preset by the system.

Optionally, in another possible implementation of the present application, "method of parsing the bitstream based on the candidate-division-mode set corresponding to the first node to determine the first information, the second information and the third information" is specifically As follows: The first numeric value is determined first. The first numeric value is the quantity of splitting modes included in the candidate-split-mode set corresponding to the first node. The bitstream is then parsed based on the first numeric value to determine the first information, second information, and third information.

The candidate-mode set corresponding to the node of the second-level coding tree may include up to 15 splitting modes. For some of the fifteen split modes, the method of obtaining the first information, the second information and the third information is the same or similar. Accordingly, the decoder device may determine the first information, the second information, and the third information based on the quantity of split modes (ie, the first numeric value) included in the candidate-mode set corresponding to the first node. In this way, the decoder device can more quickly determine node split mode information of the second-level coding tree.

Optionally, in another possible implementation of the present application, the node splitting mode corresponding to the second-level coding tree further includes quadtree splitting. In this case, the method of "parsing the bitstream to obtain node split mode information of the second-level coding tree" is as follows: The bitstream is parsed to obtain node split mode information of the second-level coding tree. The node splitting mode information of the second-level coding tree indicates a splitting mode corresponding to the first node of the second-level coding tree, and the splitting mode corresponding to the first node is a candidate-split-mode corresponding to the first node. One mode in the set, the candidate-mode set corresponding to the first node is determined according to the first preset splitting condition and the second preset splitting condition, and the second preset splitting condition is determined by the first node by quadtree splitting. It is used to indicate whether to restrict the split based on.

Optionally, in another possible implementation of the present application, the second preset partitioning condition includes at least one of the tenth preset partitioning subcondition and the eleventh preset partitioning subcondition. The tenth preset partitioning subcondition is as follows: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the corresponding node corresponds to the first node. The candidate-division-mode set does not include the quadtree splitting. The eleventh preset division subcondition is as follows: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is determined. It does not include the quadtree splitting.

Optionally, in another possible implementation of the present application, node splitting mode information of the second-level coding tree further includes fourth information, and the fourth information is used to indicate whether to split the first node based on quadtree splitting. do. Correspondingly, a method of "parsing the bitstream to obtain node splitting mode information of the second-level coding tree" is as follows: The bitstream is in accordance with the first preset splitting condition and the second preset splitting condition. Determine a candidate-division-mode set corresponding to the first node. The bitstream is parsed in the split mode corresponding to the first node to determine the first information, second information, third information and fourth information.

According to a second aspect, a decoder device is provided. The decoder apparatus includes an acquisition module, a parsing module, and a decoding and reconstruction module. The obtaining module is configured to obtain a bitstream including the image data. The parsing module: parses the bitstream obtained by the acquiring module to obtain node splitting mode information of the first-level coding tree, where the root node of the first-level coding tree is one coding tree unit (CTU). And a leaf node of the first-level coding tree is identified using the node split mode corresponding to the root node of the first-level coding tree and the node split mode information of the first-level coding tree; Parse the bitstream to obtain node split mode information of the second-level coding tree, wherein the node split mode information of the second-level coding tree indicates a split mode corresponding to the first node of the second-level coding tree. Is one mode corresponding to the first node in the candidate-division-mode set corresponding to the split mode first node, and the candidate-mode set corresponding to the first node is determined according to the first preset partitioning condition; The 1 preset splitting condition is used to indicate whether the first node is restricted from splitting in the target splitting mode, wherein the target splitting mode is at least one of horizontal binary splitting, horizontal ternary splitting, vertical binary splitting, and vertical ternary splitting. Wherein the root node of the second-level coding tree is a leaf node of the first-level coding tree; And if the splitting mode corresponding to the first node is no longer split, analyzing the bitstream to obtain encoding information of the first node, where the first node corresponds to one coding unit (CU). The decoding and reconstruction module is configured to decode and reconstruct the coding unit based on the encoding information of the first node obtained by the parsing module to obtain a picture corresponding to the picture data.

Optionally, in a possible implementation of the present application, the node splitting mode corresponding to the first-level coding tree is different from the node splitting mode corresponding to the second-level coding tree. The node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary tree splitting.

Optionally, in another possible implementation of the present application, the node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and three. Includes true tree splits.

Optionally, in another possible implementation of the present application, node splitting mode information of the second-level coding tree includes first information, second information, and third information. The first information is used to indicate whether to further split the first node, the second information is used to indicate the direction in which the first node is split, and the third information is used to indicate the mode in which the first node is split. Correspondingly, the parsing module specifically: parses the bitstream to determine, according to the first preset partitioning condition, a candidate-division-mode set corresponding to the first node; And parse the bitstream based on the candidate-division-mode set corresponding to the first node to determine first information, second information, and third information.

Optionally, in another possible implementation of the present application, the parsing module specifically determines: a first numeric value, where the first numeric value is a quantity of partitioning modes included in the candidate-division-mode set corresponding to the first node. ego; And parse the bitstream based on the first numeric value to determine first information, second information, and third information.

Optionally, in another possible implementation of the present application, the node splitting mode corresponding to the second-level coding tree further includes quadtree splitting. Correspondingly, the parsing module is specifically configured to parse the bitstream to obtain node split mode information of the second-level coding tree, wherein the node split mode information of the second-level coding tree is a second-level coding tree. Indicates a splitting mode corresponding to the first node of, wherein the splitting mode corresponding to the first node is one mode in the candidate-split-mode set corresponding to the first node, and the candidate-mode set corresponding to the first node is Determined according to the first preset split condition and the second preset split condition, the second preset split condition is used to indicate whether to limit the splitting of the first node based on the quadtree split.

Optionally, in another possible implementation of the present application, node splitting mode information of the second-level coding tree further includes fourth information, and the fourth information is used to indicate whether to split the first node based on quadtree splitting. do. Correspondingly, the parsing module specifically: parses the bitstream to determine a candidate-division-mode set corresponding to the first node according to the first preset division condition and the second preset division condition; And parsing the bitstream in a split mode corresponding to the first node to determine the first information, the second information, the third information, and the fourth information.

For a first preset dividing condition of the second aspect, reference is made to the description of the first preset dividing condition of the first aspect. Similarly, for the second preset division condition of the second aspect, refer to the description of the second preset division condition of the first aspect.

According to a third aspect, a decoder device is provided. The decoder device includes one or more processors, memory and a communication interface. Memory and communication interfaces are coupled to one or more processors. The memory is configured to store computer program code. Computer program code includes instructions. When one or more processors execute an instruction, the decoder device performs the method of decoding image data in either the first aspect or a possible implementation of the first aspect.

According to a fourth aspect, a computer readable storage medium is provided. Computer-readable storage media store instructions. When the instruction is executed at the decoder device, the decoder device may perform the method of decoding image data in either the first aspect or a possible implementation of the first aspect.

According to a fifth aspect, a computer program product comprising instructions is provided. When the computer program product is executed in the decoder device, the decoder device may perform the method of decoding image data in either the first aspect or a possible implementation of the first aspect.

In the present application, detailed descriptions of the implementations of the second, third, fourth, fifth, and second aspects, the third, fourth, and fifth aspects of the present invention are directed to the first and first aspects. See the detailed description of the implementation. Moreover, regarding the beneficial effect of implementing 2nd viewpoint, 3rd viewpoint, 4th viewpoint, 5th viewpoint and 2nd viewpoint, 3rd viewpoint, 4th viewpoint, and 5th viewpoint, implementation of a 1st viewpoint and a 1st viewpoint See the analysis of the effects of the benefits. Details are not described here again.

In the present application, the name of the decoder device described above does not constitute any limitation on the device or the functional module. In actual implementations, devices or functional modules may appear under different names. If the functionality of an apparatus or a functional module is similar to that described in this application, each apparatus or functional module is within the scope defined by the claims and equivalent techniques herein.

According to a sixth aspect, a method of decoding image data is provided. After determining the CTU corresponding to the picture block to be encoded, the encoder device divides the CTU into the node partitioning mode corresponding to the first-level coding tree to obtain a leaf node of the first-level coding tree. The root node of the first-level coding tree corresponds to the CTU. The encoder apparatus determines a candidate-division-mode set corresponding to the first node of the second-level coding tree. The candidate-division-mode set corresponding to the first node satisfies a first preset partitioning condition used to indicate whether the first node is to be restricted from splitting in the target splitting mode, and the target splitting mode is horizontal binary splitting, horizontally. And at least one of ternary division, vertical binary division, and vertical ternary division, wherein the root node of the second-level coding tree is a leaf node of the first-level coding tree. If the candidate-division-mode set corresponding to the first node includes non-additional division, the encoder device encodes the coding unit corresponding to the first node to obtain a coding unit bitstream corresponding to the coding unit.

The first preset splitting condition in this embodiment of the present application limits the splitting mode of the nodes of the second-level coding tree, greatly reducing the complexity of splitting the nodes of the second-level coding tree and reducing the coding complexity. .

Optionally, in a possible implementation of the present application, the node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary. Includes tree splitting.

Optionally, in another possible implementation of the present application, the node splitting mode corresponding to the second-level coding tree further includes quadtree splitting. In this case, the method of "determining the candidate-division-mode set corresponding to the first node of the second-level coding tree" is as follows: The candidate-division-mode set corresponding to the first node is determined. The candidate-division-mode set corresponding to the first node satisfies the first preset splitting condition and the second preset constraint, and the second preset splitting condition indicates that the first node is split based on the quadtree splitting. Used to indicate whether to limit.

Optionally, in another possible implementation of the present application, if the candidate-division-mode set corresponding to the first node includes a division mode other than non additional division, the rate distortion cost of each division mode available to the first node Is calculated; The splitting mode corresponding to the minimum rate distortion cost is determined as the target splitting mode corresponding to the first node; The first node is partitioned using the target partitioning mode corresponding to the first node.

For the first preset dividing condition in the sixth aspect, reference is made to the description regarding the first preset dividing condition in the first aspect. Similarly, for the second preset division condition of the sixth aspect, reference is made to the description regarding the second preset division condition of the first aspect.

According to a seventh aspect, an encoder device is provided. The encoder apparatus includes a decision module, a segmentation module and an encoding module. The determining module is configured to determine a coding tree unit CTU corresponding to the picture block to be encoded. The splitting module is configured to split the CTU determined by the determining module in the node splitting mode corresponding to the first-level coding tree to obtain a leaf node of the first-level coding tree. The root node of the first-level coding tree corresponds to the CTU. The determining module is also configured to determine a candidate-division-mode set corresponding to the first node of the second-level coding tree. The candidate-division-mode set corresponding to the first node satisfies the first preset dividing condition, the first preset dividing condition is used to indicate whether the first node is to be restricted from splitting in the target dividing mode, and the second The root node of the level coding tree is a leaf node of the first-level coding tree, and the target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. The encoding module: if the candidate-division-mode set corresponding to the first node and determined by the determining module includes non-additional division, encoding the coding unit corresponding to the coding unit by encoding a coding unit (CU) corresponding to the first node- Configured to obtain a unit bitstream.

Optionally, in a possible implementation of the present application, the node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary. Includes tree splitting. .

Optionally, in a possible implementation of the present application, the node splitting mode corresponding to the second-level coding tree further includes quadtree splitting. Correspondingly, the determining module is specifically configured to determine a candidate-division-mode set corresponding to the first node. The candidate-division-mode set corresponding to the first node satisfies the first preset splitting condition and the second preset constraint, and the second preset splitting condition indicates that the first node is split based on the quadtree splitting. Used to indicate whether to limit.

Optionally, in a possible implementation of the present application, the encoder device provided by this embodiment of the present application further includes a calculation module. The calculating module calculates the rate distortion cost of each partitioning mode available for the first node if the candidate-divisioning-mode set corresponding to the first node and determined by the determining module includes a partitioning mode other than non-additional partitioning. Configured to calculate. Correspondingly, the determining module is further configured to determine the splitting mode corresponding to the minimum rate distortion cost as the target splitting mode corresponding to the first node. The splitting module is specifically configured to split the first node in the target splitting mode corresponding to the first node and determined by the determining module.

For the first preset dividing condition in the seventh aspect, reference is made to the description regarding the first preset dividing condition in the first aspect. Similarly, for the second preset dividing condition of the seventh aspect, reference is made to the description regarding the second preset dividing condition of the first aspect.

According to an eighth aspect, an encoder device is provided. The encoder device includes one or more processors, memory and communication interface. Memory and communication interfaces are coupled to one or more processors. The memory is configured to store computer program code. Computer program code includes instructions. When one or more processors execute an instruction, the encoder apparatus performs the method of encoding image data in either the sixth aspect or a possible implementation of the sixth aspect.

According to a ninth aspect, a computer readable storage medium is provided. Computer-readable storage media store instructions. When the instruction is executed in the encoder device, the encoder device may perform the method of encoding image data in either the sixth aspect or a possible implementation of the sixth aspect.

According to a tenth aspect, a computer program product comprising instructions is provided. When the computer program product is executed in the encoder device, the encoder device may perform the method of encoding picture data in either the sixth aspect or a possible implementation of the sixth aspect.

Detailed descriptions of the implementation of the seventh, eighth, ninth, tenth, and seventh, eighth, ninth, and tenth aspects are detailed descriptions of the implementation of the seventh and seventh aspects. See. In addition, regarding the beneficial effects of the implementation of the seventh, eighth, ninth, tenth and seventh, eighth, ninth, and tenth aspects, the sixth and sixth aspects are implemented. See the analysis of the beneficial effects of. Details are not described here again.

In the present application, the name of the encoder device does not constitute any limitation on the device or the functional module. In actual implementations, devices or functional modules may appear under different names. If the functionality of an apparatus or a functional module is similar to that described in this application, each apparatus or functional module is within the scope defined by the claims and equivalent techniques herein.

These or other aspects in the present application are more concise and understandable in the following description.

1 is a schematic structural diagram of different partitioning modes according to an embodiment of the present application.
2 is a schematic structural diagram of a depth of a quadtree according to an embodiment of the present application.
3 is a schematic diagram of combined partitioning of quadtree partitioning and binary tree partitioning according to an embodiment of the present application.
4 is a schematic structural diagram of an image processing system according to an embodiment of the present application.
5 is a schematic diagram of a hardware structure of a mobile telephone according to an embodiment of the present application.
6 is a first schematic flowchart of a method of decoding image data according to an embodiment of the present application.
7 is a second schematic flowchart of a method of decoding image data according to an embodiment of the present application.
8 is a schematic flowchart of an image data encoding method according to an embodiment of the present application.
9 is a schematic structural diagram of a decoder device according to an embodiment of the present application.
10 is a schematic structural diagram of an encoder device according to an embodiment of the present application.

In the specification, claims and appended drawings of this application, terms such as "first", "second", "third" and "fourth" are intended to distinguish different objects but do not indicate a particular order.

In the embodiments of the present application, words such as "yes" or "for example" are used to provide an example, description or description. Any embodiment or design manner described as "example" or "for example" in the embodiments of the present application should not be construed as having the desired or more advantages over other embodiments or design approaches. Exactly, words such as "yes" or "for example" are used to present related concepts in a particular way.

In order to facilitate understanding of the embodiments of the present application, the relevant elements in the embodiments of the present application are described first.

Coding Tree Unit (CTU): A picture includes a plurality of CTUs. CTU generally corresponds to one square plot area. As shown in FIG. 1A, the image 10 includes a plurality of CTUs (including CTU A, CTU B, CTU C, etc.). The encoding information corresponding to the CTU includes luminance values and / or chrominance values of the pixels in the square image area corresponding to the CTU. In addition, the encoding information corresponding to the CTU may further include a syntax element. The syntax element represents a method of dividing a CTU into at least one CU and decoding each CU to obtain a reconstructed picture.

The picture area corresponding to one CTU may include 64 x 64, 128 x 128 or 256 x 256 pixels. In an example, a CTU comprising 64 x 64 pixels includes a rectangular pixel matrix having 64 columns and 64 rows of pixels, each pixel comprising a luminance component and / or a chrominance component.

It should be understood that a CTU may alternatively correspond to a rectangular picture area or another shape of picture area, and that a picture area corresponding to one CTU may alternatively be a picture area with a horizontal number of pixels. The direction differs from the number of pixels in the vertical direction and includes, for example, 64 x 128 pixels.

Coding Unit (CU): A coding unit generally corresponds to one AxB rectangular area, where A is the width of the rectangle and B is the height of the rectangle. The width in the embodiment of the present application is the length in the X axis direction (horizontal direction) in the two-dimensional rectangular coordinate system XoY shown in FIG. 1. The height in the two-dimensional rectangular coordinate system XoY shown in FIG. 1 is the length in the Y axis direction (vertical direction). Here, the values of A and B may be the same or different. The values of A and B are generally integer powers of two, such as 256, 128, 64, 32, 16, 8 or 4. The reconstructed picture block of the AxB rectangular area can be obtained through the decoding process of the CU. Decoding processes generally include processing such as prediction, dequantization, and inverse transform, where predicted pictures and residuals are generated, and predicted pictures and residuals are superimposed and reconstructed picture blocks. Acquire. The final reconstructed picture can be obtained using a plurality of reconstructed picture blocks.

Quad-Tree (QT): A quadtree is a tree structure, and nodes may be divided into four child nodes. For example, in the video codec standard H.265, quadtree-based CTU splitting mode is used. CTU is used as the root node. Each node corresponds to one square picture area. The node may not be further divided (in this case, the square picture area corresponding to the node is a CU); Or divided into four lower level nodes, that is, the square image area corresponding to the node is divided into four identical square areas (the width and height of the square area are each half width and half height of the pre-division area), Each zone corresponds to one node. As shown in Fig. 1 (b), CTU A is the root node, and CTU A is divided into four nodes a, b, c and d. If node a is non-additional division, the square picture area corresponding to node a corresponds to one CU.

Binary Tree (BT): A binary tree is a tree-like structure. The node may be divided into two child nodes. Nodes in the binary tree structure may no longer be split or may be split into two lower-level nodes. Binary tree splitting mode may include one of the following.

(1) horizontal binary division

The picture area corresponding to the node is divided into an upper area and a lower area having the same size. Specifically, for the image area obtained through division, the width does not change and the height is half of the divided image area. Each picture area obtained through the division corresponds to one child node. As shown in Fig. 1 (c), node b is partitioned in horizontal binary tree splitting mode to produce node e and node f.

(2) vertical binary division

The area corresponding to the node is divided into a left area and a right area having the same size. Specifically, in the image area obtained through division, the height does not change and the width is half of the divided image area. Each picture area obtained through the division corresponds to one child node. As shown in Fig. 1 (d), node d is split in vertical binary division mode to produce node g and node h.

Horizontal binary division and vertical binary division can be understood as examples of binary tree division modes. The picture area corresponding to the node may be divided into two sub areas in different modes. For example, you can get two subregions with different heights through horizontal divisions, or two regions with different widths through vertical divisions.

Ternary Tree (TT): A ternary tree is a tree-like structure. The node can be divided into three child nodes. Nodes in the ternary tree structure may no longer be split or may be split into three lower level nodes. The ternary tree splitting mode may include one of the following.

(1) horizontal ternary division

The image area corresponding to the node is divided into three areas, namely, an upper area, an intermediate area and a lower area, each area corresponding to one child node. In one example, the heights of the upper region, the middle region and the lower region are respectively 1/4, 1/2 and 1/4 of the height of the image region before dividing. As shown in Fig. 1 (e), node c is partitioned in the horizontal ternary partitioning mode to generate node j, node k and node m. The height of the image area corresponding to node j is 1/4 of the height of the image area corresponding to node c, the height of the image area corresponding to node k is 1/2 of the height of the image area corresponding to node c, and node m The height of the image area corresponding to is 1/4 of the height of the image area corresponding to node c. In another example, the heights of the upper region, the middle region and the lower region are 1/3, 1/3 and 1/3 of the height of the image region before dividing, respectively. Specifically, the image region corresponding to the node is equally divided into three regions in the X-axis direction in the two-dimensional rectangular coordinate system XoY shown in FIG. As shown in Fig. 1 (f), node c is partitioned in the horizontal ternary partitioning mode to generate node j, node k and node m; The height of the image region corresponding to the node j, the height of the image region corresponding to the node k, and the height of the image region corresponding to the node m are all 1/3 of the height of the image region corresponding to the node c.

(2) vertical ternary division

The image area corresponding to the node is divided into three areas, that is, the left area, the middle area and the right area, each area corresponding to one child node. In one example, the widths of the left region, the middle region and the right region are 1/4, 1/2 and 1/4 of the width of the image region before dividing, respectively. As shown in Fig. 1 (g), node c is divided into vertical ternary partitioning modes to produce node p, node q and node x. The width of the image region corresponding to the node p is 1/4 of the width of the image region corresponding to the node c, the width of the image region corresponding to the node q is 1/2 the width of the image region corresponding to the node c, The width of the image area corresponding to node x is 1/4 of the width of the image area corresponding to node c. In another example, the heights of the left region, the middle region and the right region are 1/3, 1/3 and 1/3 of the height of the image region before dividing, respectively. Specifically, the image region corresponding to the node is equally divided into three regions in the X-axis direction in the two-dimensional rectangular coordinate system XoY shown in FIG. As shown in Fig. 1 (h), node c is partitioned in the vertical ternary partitioning mode to generate node p, node q and node x; The width of the image region corresponding to the node p, the width of the image region corresponding to the node q, and the width of the image region corresponding to the node x are all 1/3 of the width of the image region corresponding to the node c.

Image encoding: Image encoding is a processing process in which a picture sequence is compressed into a bitstream.

Image Decoding (Image Decoding): Image decoding is a processing process in which a bitstream is reconstructed into a picture reconstructed according to a specific parsing rule and a specific processing method.

In the video codec standard H.265, the CTU is split into QT split mode. In particular, the CTU is used as the root node of the QT structure, and the CTU is recursively divided into several leaf nodes in the QT partitioning mode. When a node is no longer split, the node is called a leaf node. It can be seen from the above description that an image includes a plurality of CTUs, and one CTU corresponds to one square image area. That is, one CTU corresponds to one picture block. Each leaf node corresponds to one CU. Each CU is equivalent to a child picture block in the picture block corresponding to the CTU, and the child picture block can no longer be split in the QT split mode. If the node needs to be further divided, the picture area corresponding to the node is divided into four picture areas of the same size. Referring to Fig. 1 (b), after the image region corresponding to the node is divided, each image region generated after the division corresponds to one node, and it must be further determined whether these nodes are further divided. Whether to split the node is identified by a split flag bit (eg, split_cu_flag) corresponding to the node in the bitstream. In the QT structure, the depth of the root node (abbreviated as QT depth) is 0, and the QT depth of the child node generated in the QT partitioning mode is 1 plus the QT depth of the parent node of the child node.

For example, the split flag bit of the node is indicated by split_cu_flag. split_cu_flag = 0 indicates that the node is no longer split, or split_cu_flag = 1 indicates that the node is split further. As shown in FIG. 2, if the value of split_cu_flag of a 64x64 CTU node (with a QT depth of 0) is 1, the CTU node is split into four 32x32 nodes (with a QT depth of 1), and four 32x32 nodes each. Nodes A1, node A2, node A3 and node A4. Each of the four 32x32 nodes may or may not be split based on the split_cu_flag corresponding to the node. If the value of split_cu_flag of node A1 is 1, node A1 is further divided into four 16x16 nodes (QT depth 2), and the four 16x16 nodes are node B1, node B2, node B3, and node B4, respectively. Continue until no longer split. In this way, one CTU is divided into CU groups. In the same QT structure, the minimum size of all CUs (ie, the minimum value of a CU) is the same, and the minimum size of the CU is identified in a sequence parameter set (SPS) of the bitstream. For example, an 8x8 CU is the minimum CU. In the above recursive partitioning process, if the size of the node is equal to the minimum size of the CU, it is determined that the node is no longer split and the splitting flag of the node need not be included in the bitstream.

The splitting mode in which the CTU is divided into groups of CUs corresponds to one coding tree. 3 (a) is an example of a coding tree. The coding tree may correspond to one splitting mode, eg, quadtree splitting mode, or may correspond to a plurality of splitting modes, for example, the above-described QTBT structure and the below-described QT-BT / TT structure.

The first-level coding tree, the second-level coding tree, the third-level coding tree, ... and the N-th level coding tree correspond to different split mode sets, where N is the different split mode set, the type of tree, For example, it can be a ternary tree, binary tree or quadtree; Or a split mode set in the same type of tree, eg, horizontal binary division or vertical binary division; Or combinations thereof. It can be appreciated that the coding tree does not necessarily need to include all the plurality of coding trees described above at different levels. For example, the coding tree may include only the first-level coding tree; Or the coding tree may comprise a first-level coding tree and a second-level coding tree; Alternatively, the coding tree may include a first-level coding tree, a second-level coding tree and a third-level coding tree.

In one example, the first-level coding tree can include quadtree splitting, and the second-level coding tree can include binary tree splitting and ternary tree splitting.

In one example, the first-level coding tree can include quadtree splitting, and the second-level coding tree can include binary tree splitting, ternary tree splitting, and quadtree splitting.

In one example, the first-level coding tree may comprise quadtree splitting and binary tree splitting, and the second-level coding tree comprises ternary tree splitting.

In one example, the first-level coding tree can include quadtree splitting, the second-level coding tree can include binary tree splitting, and the third-level coding tree can include ternary tree splitting.

In one example, the first-level coding tree may comprise a horizontal binary division, the second-level coding tree may comprise a vertical binary division and a quadtree division, and the third-level coding tree may comprise a vertical ternary division and It may include a horizontal ternary split.

The second-level coding tree may further include other splitting modes, which are not particularly limited in the embodiments of the present application.

The encoder device generally uses Rate Distortion Optimization (RDO) rate technology to determine the particular coding tree to use in the CTU for encoding. Specifically, for each node, the encoder device calculates the rate distortion cost (RD cost) of each split mode that can be used by the node, compares the calculated RD cost, and corresponds to the minimum RD cost. The split mode is determined as the split mode of the node.

Similar to the QT structure, the depth of the node in the BT structure is referred to as the BT depth, and the BT depth of the child node created in the BT split mode is the BT depth of the parent node. If the node's BT depth equals the maximum BT depth, it is determined that the node is no longer split. The maximum BT depth of the BT structure is usually identified in the SPS.

For example, the BT splitting mode is introduced in addition to the QT splitting mode. QT split mode and BT split mode are cascaded to obtain split mode. This split mode is called QTBT split mode. Specifically, the CTU may be split in the QT splitting mode, and the QT leaf node may be split further in the BT splitting mode. In other words, the first-level coding tree is QT and the second-level coding tree is BT.

As shown in Fig. 3 (a), each endpoint represents a node, a solid line indicates a node is split in QT split mode, a dotted line indicates a node is split in BT split mode, and A to M are 13 leaves It is a node, and each leaf node corresponds to one CU. In the BT structure, 10 represents vertical binary division and 11 represents horizontal binary division. FIG. 3 (b) shows an image area corresponding to the divided CTU in the division mode shown in FIG. 3 (a).

In QTBT split mode, each CU has a QT depth and a BT depth. In a CU of QTBT split mode, the QT depth of the CU is the QT depth of the QT leaf node to which the CU belongs, and the BT depth of the CU is the BT depth of the BT leaf node to which the CU belongs. If the CTU is not split, i.e. there is only one CU, the QT depth of the CU is zero and the BT depth of the CU is zero.

For example, in FIG. 3A, the QT depths of A and B are 1 and the BT depths of A and B are 2. The QT depths of C, D and E are 1, and the BT depths of C, D and E are 1; The QT depths of F, K and L are 2, and the BT depths of F, K and L are 1; The QT depths of I and J are 2, and the BT depths of I and J are 0; The QT depths of G and H are 2, the BT depth of G and H is 2, the QT depth of M is 1, and the BT depth of M is 0.

The CU shape of the QTBT splitting mode is more diversified, which can better adapt to the content of the local picture. In standard H.265, all CUs obtained through splitting based on the QT splitting mode may be rectangular, that is, the width of the CU may be equal to the height of the CU. After the BT splitting mode is introduced in addition to the QT splitting mode, the width and height of the CU may be different. For example, the width-to-height ratio (the numeric value of the width-to-height ratio equals the width divided by the height) is 2, 4, 8, 16, 1/2, 1/4, 1/8, or 1/16. to be. Certainly, in the QTBT splitting mode, the width and height of each CU cannot be smaller than the side length of the minimum CU (eg, the minimum CU can be set to 4 × 4). The SPS of the video stream generally contains the size information of the minimum CU.

In addition to the QTBT, a QT-BT / TT splitting mode may be further included. Specifically, nodes of the first-level coding tree may use the QT splitting mode, and nodes of the second-level coding tree may use the BT splitting mode or the TT splitting mode. Specifically, the CTU is the root node of the first-level coding tree, and the CTU is split in QT splitting mode to generate several leaf nodes of the first-level coding tree. Then, the leaf nodes of the first-level coding tree are used as root nodes of the second-level coding tree, and the nodes of the second-level coding tree are BT splitting modes (including horizontal binary splitting and vertical binary splitting). Or split in TT splitting mode (including horizontal ternary splitting and vertical ternary splitting) to create several leaf nodes of the second-level coding tree.

However, for each node, the encoder device generally calculates the RD cost of each splitting mode that can be used by the node, compares the calculated RD costs, and splits the splitting mode corresponding to the minimum RD cost. Determine the mode. In the QT-BT / TT splitting mode, since the BT splitting mode or the TT splitting mode can be used for the nodes of the second-level coding tree, for each node of the second-level coding tree, the encoder apparatus is divided into four splitting modes. RD costs of (horizontal binary division, vertical binary division, horizontal ternary division and vertical ternary division) are calculated to determine the partition mode actually used by the nodes of the second-level coding tree. As a result, the encoding complexity is relatively high.

In view of the above problem, an embodiment of the present application provides a method of decoding image data. After acquiring the bitstream including the image data, the decoder device parses the bitstream to obtain node dividing mode information of the first-level coding tree and node dividing mode information of the second-level coding tree. The node splitting mode information of the second-level coding tree indicates a splitting mode corresponding to the first node of the second-level coding tree, and the splitting mode corresponding to the first node is a candidate-split-mode corresponding to the first node. One mode in the set, and the candidate-mode set corresponding to the first node is determined according to the first preset splitting condition, wherein the first preset splitting condition indicates whether to limit the first node to split in the target splitting mode. The target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. If the splitting mode corresponding to the first node is non-additional splitting, the decoder device parses the bitstream to obtain encoding information of the first node. The first node corresponds to one coding unit CU. In this way, the decoder device can decode and reconstruct the coding unit based on the encoding information of the first node to obtain a picture corresponding to the picture data. In the present application, the root node of the first-level coding tree corresponds to the CTU, and the root node of the first-level coding tree is the root node of the first-level coding tree and the node split mode information of the first-level coding tree. The root node of the second-level coding tree is identified using the node splitting mode corresponding to the leaf node of the first-level coding tree. In this embodiment of the present application, the splitting mode corresponding to the first node of the second-level coding tree is one of splitting modes determined according to the first preset splitting condition, and the first preset splitting condition is determined by the first node. Used to indicate whether to limit the division in the target division mode, the target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. The first preset partitioning condition reduces the quantity of partitioning modes corresponding to the first node, effectively reducing the decoding complexity.

Correspondingly, an embodiment of the present application further provides a method of encoding image data. After determining the CTU corresponding to the picture block to be encoded, the encoder device divides the CTU into the node partitioning mode corresponding to the first-level coding tree to obtain a leaf node of the first-level coding tree. The root node of the first-level coding tree corresponds to the CTU. The encoder apparatus determines a candidate-division-mode set corresponding to the first node of the second-level coding tree. The candidate-division-mode set corresponding to the first node satisfies a first preset partitioning condition used to indicate whether the first node is to be restricted from splitting in the target splitting mode, and the target splitting mode is horizontal binary splitting, horizontally. And at least one of ternary division, vertical binary division, and vertical ternary division, wherein the root node of the second-level coding tree is a leaf node of the first-level coding tree. If the candidate-division-mode set corresponding to the first node includes non-additional division, the encoder device encodes the coding unit corresponding to the first node to obtain a coding unit bitstream corresponding to the coding unit. The first preset splitting condition in this embodiment of the present application limits the splitting mode of the nodes of the second-level coding tree, greatly reducing the complexity of splitting the nodes of the second-level coding tree and reducing the coding complexity. Let's do it.

The image data encoding method and the image data decoding method provided by the embodiments of the present application are applicable to both an image processing system. 4 is a schematic structural diagram of an image processing system according to an embodiment of the present application. As shown in FIG. 4, the image processing system includes an encoder device 40 and a decoder device 41. The encoder device 40 and the decoder device 41 may be individually arranged or integrated into one device. This is not particularly limited in this embodiment of the present application. For example, in FIG. 1, the encoder device 40 and the decoder device 41 are arranged separately. For convenience of description, the description is provided below using an example in which the encoder device and the decoder device are arranged separately.

Specifically, after capturing the video, the encoder device 40 processes the CTU corresponding to each picture of the video in the split mode of the first-level coding tree and the split mode of the second-level coding tree. If the candidate-division-mode set corresponding to the first node of the second-level coding tree includes non-additional partitioning, the encoder device 40 encodes the coding unit corresponding to the first node to code the corresponding coding unit. Acquire a unit bitstream. After obtaining the coding unit bitstream corresponding to each node of the second-level coding tree, the encoder device 40 obtains the CTU bitstream and sends the CTU bitstream to the decoder device 41. The decoder device 41 obtains the node split mode information of the second-level coding tree, and corresponds to the decoder device 41 based on the node split mode information of the second-level coding tree to obtain a reconstructed picture. Perform the process.

Encoder device 40 and decoder device 41 may be various devices having a camera (eg, a front camera or a rear camera). For example, the encoder device and decoder device may be a wearable electronic device (eg a smart watch) or an instant camera; The mobile phone shown in FIG. 5 may be a tablet computer, a desktop computer, a virtual reality device, a notebook computer, an ultra-mobile personal computer (UMPC), a personal digital assistant (PDA), or the like. have. The specific form of the encoder device 40 and the decoder device 41 is not particularly limited in this embodiment of the present application.

Referring to FIG. 4, as shown in FIG. 5, in this embodiment, both the encoder device 40 and the decoder device 41 may be mobile phones. Hereinafter, the present embodiment will be described in detail using a mobile phone as an example.

The mobile phone shown in FIG. 5 is only one example of the encoder device 40 and the decoder device 41, and the mobile phone may include more or less components than those shown in FIG. 5, and two or more components. It is to be understood that the combinations or the configurations may have different component configurations. The components shown in FIG. 5 may be implemented in hardware, including one or more signal processors and / or application specific integrated circuits, software, or a combination of hardware and software.

As shown in FIG. 5, the mobile telephone includes a radio frequency (RF) circuit 50, a memory 51, an input unit 52, a display unit 53, a sensor 54, and an audio circuit 55. ), A wireless fidelity (Wi-Fi) module 56, a processor 57, a Bluetooth module 58, and a power supply 59. Those skilled in the art will appreciate that the structure of the mobile telephone shown in Figure 5 is a mobile telephone. It is to be understood that a mobile telephone may include more or fewer components than shown in FIG. 5, and may combine some components or have a different component arrangement than does not constitute a limitation to.

Next, the components of the mobile telephone will be described in detail with reference to FIG. 5.

The RF circuit 50 may be configured to transmit and receive signals in an information transceiving process or a calling process. The RF circuit 310 may receive the downlink information from the base station and then transmit the downlink information to the processor 57 for processing; And transmit uplink data to the base station. RF circuitry generally includes, but is not limited to, devices such as antennas, at least one amplifier, transceivers, couplers, low noise amplifiers, and duplexers. In addition, the RF circuit 50 may further communicate with the network and other mobile devices via wireless communication. Wireless communications may use any communication standard or protocol, including but not limited to global systems for mobile communications, general packet radio services, code division multiple access, broadband code division multiple access, long-term evolution, email, short message services, and the like. Can be.

The memory 51 may be configured to store software programs and data. The processor 57 executes data stored in the software program and the memory 51 to perform various functions of the mobile phone 110 and process the data. The memory 51 may mainly include a program storage area and a data storage area. The program storage area may store an operating system, an application program required for at least one function (for example, an audio reproduction function and a photo reproduction function). The data storage area can store data (audio data, phone book and video, etc.) generated according to the use of the mobile phone. In addition, the memory 51 may include fast random access memory and may further include nonvolatile memory, such as at least one magnetic disk storage device, flash memory device or other volatile solid state storage device. In the following embodiments, the memory 51 is an operating system capable of running a mobile phone, such as an iOS® operating system developed by Apple, an Android® open source operating system developed by Google, and a Windows® operating system developed by Microsoft. Save your system.

The input unit 52 (eg, touch screen) may be configured to receive input numeric or text information and to generate signal inputs related to user settings and function control of the mobile phone. In detail, the input unit 52 may include a touch screen 521 and another input device 522. Touch screen 521, also referred to as touch screen, is a touch operation performed by a user on or near touch screen 521 (e.g., the user touches the touch screen using any suitable object or accessory, such as a finger or a stylus). Operation performed on or near 521) to drive the corresponding connection device according to a preset program. Optionally, the touch screen 521 may include two components, a touch detection device and a touch controller (not shown in FIG. 5). The touch detection device detects a touch position of a user, detects a signal generated by a touch operation, and transfers the signal to the touch controller. The touch controller may receive touch information from the touch detection device, convert the touch information into contact coordinates, transmit the contact coordinates to the processor 57, and receive and execute a command sent by the processor 57. In addition, the touch screen 521 may be implemented in various types such as a resistance type, a capacitance type, an infrared type, and a surface acoustic wave type.

The display unit 53 (ie, display screen) may be configured to display information input by the user or information provided to the user, and a graphical user interface (GUI) of various menus of the mobile phone. The display unit 53 may include a display panel 531 disposed on the front of the mobile phone. Alternatively, the display panel 531 may be configured in the form of a liquid crystal display (LCD) or an organic light-emitting diode (OLED). In addition, the touch screen 521 may cover the display panel 531. After detecting the touch action on or near the touch screen 521, the touch screen 131 transfers the touch action to the processor 57 to determine the touch event type. The processor 57 then provides a corresponding visual output on the display panel 531 based on the touch event type. In FIG. 5, the touch screen 521 and the display panel 531 are used as separate components to implement input and output functions of the mobile phone. However, in some embodiments, touch screen 521 and display panel 531 may be integrated to implement the input and output functions of a mobile phone.

In other embodiments, a pressure sensor may also be configured on the touch screen 521. In this way, when the user performs a touch operation on the touch screen, the touch screen can further detect the pressure caused by the touch operation, so that the mobile phone can detect the touch operation more accurately.

The mobile phone may further comprise at least one sensor 54, for example an optical sensor, a motion sensor and another sensor. In detail, the optical sensor may include an ambient light sensor and a proximity sensor. The ambient light sensor may adjust the brightness of the display panel 531 based on the brightness of the ambient light. Proximity sensors are placed in front of the mobile phone. When the mobile phone moves to the ear, the mobile phone turns off the display panel 531 based on the detection of the proximity sensor, and the mobile phone can save more power. As a motion sensor, the accelerometer sensor can detect the magnitude of acceleration in all directions (usually three axes), detect gravity and direction when the accelerometer sensor is stationary, and can be used for mobile phone posture applications (e.g., landscapes). Screen switching between mode and portrait mode, associated game and magnetometer posture correction, and functions associated with vibration recognition (eg, pedometer or knock). Other sensors such as gyroscopes, barometers, hygrometers, thermometers and infrared sensors may be further configured in the mobile phone. Details are not described here.

Audio circuitry 55, speaker 551, and microphone 552 may provide an audio interface between a user and a mobile phone. The audio circuit 55 may transmit an electrical signal converted from the received audio data to the speaker 551, which converts the electrical signal into a sound signal for output. In addition, the microphone 552 converts the collected sound signal into an electrical signal, and the audio circuit 55 receives the electrical signal and then converts the electrical signal into audio data, and then converts the audio data into the RF circuit 50. The audio data is sent to another mobile phone, for example, or the audio data is output to the memory 51 for further processing.

Wi-Fi is a near field transmission technology. By using the Wi-Fi module 56, the mobile phone can help the user to send and receive email, browse web pages, access streaming media, and the like. Wi-Fi provides users with wireless broadband Internet access.

Processor 57 is the control center of mobile phone 110 and is connected to the components of the mobile phone as a whole using various interfaces and lines. The processor 57 executes or executes a software program stored in the memory 51 and calls data stored in the memory 51 to perform various functions of the mobile phone and process the data to perform overall monitoring of the mobile phone. do. In some embodiments, processor 57 may include one or more processing units. The processor 57 may further integrate an application processor and a modem processor. Application processors mainly handle operating systems, user interfaces, and applications. The modem processor mainly handles wireless communications. It is to be understood that the modem processor may also be deployed separately.

The Bluetooth module 58 is configured to exchange information with another device using a short range communication protocol such as Bluetooth. For example, the mobile phone can use the Bluetooth module 58 to establish a Bluetooth connection to a wearable electronic device (eg, smart watch) having a Bluetooth module to exchange data.

The mobile phone further includes a power source 59 (eg, a battery) that powers the component. Power can be logically connected to the processor 57 using a power management system, so that functions such as charge and discharge management and power consumption management can be implemented using the power management system.

With reference to a specific embodiment, the following describes in detail the image data encoding method and the image data decoding method according to an embodiment of the present application.

6 is a schematic flowchart of an image data decoding method according to an embodiment of the present application. The decoding method can be applied to the image processing system shown in FIG.

As shown in Fig. 6, the image data decoding method includes the following steps.

S600. The decoder device obtains a bitstream containing the image data.

Optionally, the bitstream including the picture data and obtained by the decoder device may be a sequence parameter set (SPS), picture parameter set (PPS), slice header or slice segment header ( slice segment header) and a CTU bitstream. The CTU bitstream carries picture data.

S601. The decoder device decodes the bitstream obtained by the decoder device to obtain node division mode information of the first-level coding tree.

The root node of the first-level coding tree corresponds to one CTU, and the leaf nodes of the first-level coding tree are identified using the node split mode corresponding to the root node and node splitting of the first-level coding tree. . Mode information of the first-level coding tree. The node splitting mode corresponding to the first-level coding tree includes quadtree splitting.

After obtaining the bitstream, the decoder device parses the CTU bitstream from the bitstream to obtain node partitioning information of the first-level coding tree.

Optionally, a method for parsing a CTU bitstream by a decoder device to obtain node partitioning information of the first-level coding tree may be as follows: The decoder device may use the CTU bit by using the CTU as the first root node. Obtain a first identifier (eg, SplitFlag) that is included in the syntax element of the stream and used to indicate how to split the CTU into at least one CU, i.e., the first identifier is a node of the first-level coding tree. Indicates split mode information. In an example, if the numeric value of the particular SplitFlag is 0, this indicates that the node corresponding to the particular SplitFlag is a leaf node of the first-level coding tree. If the numerical value of SplitFlag is 1, SplitFlag of four child nodes of the node corresponding to the specific SplitFlag is further obtained until information on all leaf nodes of the first-level coding tree is determined.

If the width of the picture area corresponding to the node is equal to the first threshold (eg, if the first threshold is 8 or 16), then the node is a leaf node of the first-level coding tree, and the number of SplitFlags corresponding to the node. The value is zero.

Optionally, the method for parsing the CTU bitstream by the decoder device to obtain node partitioning information of the first-level coding tree may also be as follows: The decoder device may determine the CTU as the root node of the first-level coding tree. Parsing the CTU bitstream to obtain a second identifier (eg, NSFlag) used to indicate whether to split a node of the first-level coding tree while being included in the syntax element of the CTU bitstream. If the numeric value of the second identifier is a first value (eg, 1), this indicates that the node corresponding to the second identifier is a leaf node of the first-level coding tree and also a leaf node of the second-level coding tree. Or; Or if the numeric value of the second identifier is a second value (eg, 0), a third identifier (eg, QTSplitFlag) included in the syntax element is obtained. If the numeric value of the third identifier is a third value (eg, 0), it indicates that the node corresponding to the third identifier is a leaf node of the first-level coding tree but not a leaf node of the second-level coding tree. Or; Or if the numeric value of the third identifier is a fourth value (eg, 1), the four child nodes of the node corresponding to the third identifier until information about all leaf nodes of the first-level coding tree is determined. The second identifier is further obtained.

If the width of the picture area corresponding to the node is equal to the first threshold (eg, if the first threshold is 8 or 16), then the node is a leaf node of the first-level coding tree and a third identifier corresponding to the node. Is a third value. Further, if the width or height of the image area corresponding to the node is greater than the second threshold and the value of the second identifier corresponding to the node is the second value, the numeric value of the third identifier is the third value; Or if the splitting mode corresponding to the node is QT splitting, the numeric value of the second identifier corresponding to the node is the second value, and the numeric value of the third identifier corresponding to the node is the third value.

In one example, the syntax table of the first-level coding tree in this embodiment of the present application is shown in Table 1. In Table 1, coding_quadtree () is a syntax structure of the first-level coding tree and describes information about nodes of the first-level coding tree.

In Table 1, the coding_quadtree (xCtb, yCtb, CtbLog2SizeY, 0) is parsed using CTU as the root node, where CtbLog2SizeY is the log of the side length of the CTU based on 2 (where the image area corresponding to the CTU is square XCtb and yCtb represent horizontal coordinate offsets and vertical coordinate offsets of the upper left corner of the image area corresponding to the CTU, respectively, with respect to the upper left corner of the image area corresponding to the node; x0 and y0 respectively indicate a horizontal coordinate offset and a vertical coordinate offset of the upper left corner of the image area corresponding to the node with respect to the upper left corner of the image area corresponding to the CTU; log2CbSize represents the log of the side length of the picture area corresponding to the node based on 2 (the picture area corresponding to the CTU is square and the first-level coding tree uses only quadtree splitting, The image area corresponding to all nodes is also rectangular and has an equivalent width and height, only the side lengths need to be identified, and there is no need to distinguish the width and height); cqtDepth represents the depth of the node in the first-level coding tree; The condition condA indicates a condition under which the node split information syntax element split_cu_flag of the first-level coding tree needs to be parsed from the bitstream. For example, condA is "the picture area corresponding to the node is in the picture area corresponding to the CTU, and the side length of the picture area corresponding to the node is larger than the threshold value." If split_cu_flag is 0, this indicates that the node is not split based on the quadtree and is a leaf node of the first level coding tree. In this case, the leaf node is parsed based on the syntax structure coding_second_tree () of the second-level coding tree; Alternatively, if split_cu_flag is 1, it indicates that the node is split based on the quadtree. In this case, the node is divided into four child nodes, the side lengths and coordinates of the image region corresponding to each child node, and the depth of each child node in the first-level coding tree are determined, and the child node is coding_quadtree (). Are parsed sequentially. "X >> Y" indicates moving X to the right by Y bits. "X << Y" indicates that X is shifted left by Y bits. ae (v) represents parsing a syntax element based on CABAC.

coding_quadtree (x0, y0, log2CbSize, cqtDepth) { Descriptor Descriptor  if (condA) split_cu_flag [x0] [y0] ae (v) if (split_cu_flag [x0] [y0]) { x1 = x0 + (1 << (log2CbSize -1)) y1 = y0 + (1 << (log2CbSize -1)) coding_quadtree (x0, y0, log2CbSize1, cqtDepth +1) if (x1 <pic_width_in_luma_samples) coding_quadtree (x1, y0, log2CbSize1, cqtDepth +1) if (y1 <pic_height_in_luma_samples) coding_quadtree (x0, y1, log2CbSize1, cqtDepth +1) if (x1 <pic_width_in_luma_sample && y1 <pic_height_in_luma_samples) coding_quadtree (x1, y1, log2CbSize1, cqtDepth +1) } else coding_second_tree (x0, y0, log2CbSize, log2CbSize, 0, cqtDepth) }

S602. The decoder device parses the bitstream obtained by the decoder device to obtain node division mode information of the second-level coding tree.

Node splitting mode information of the second-level coding tree indicates a splitting mode corresponding to the first node of the second-level coding tree, and a splitting mode corresponding to the first node is a candidate-split-mode corresponding to the first node. One mode in the set, the candidate-division-mode set corresponding to the first node is determined according to the first preset division condition, and the first preset division condition restricts the first node from being split in the target division mode. Used to indicate whether or not to split, the target split mode is one of horizontal binary split, horizontal ternary split, vertical binary split, and vertical ternary split. At least one.

The root node of the second-level coding tree is a leaf node of the first-level coding tree. The node splitting mode corresponding to the second-level coding tree is different from the node splitting mode corresponding to the first-level coding tree. In this embodiment, the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary tree splitting, the binary tree splitting comprises horizontal binary splitting and vertical binary splitting, Binary tree splitting includes horizontal ternary splitting and vertical ternary splitting.

After acquiring node splitting mode information of the first-level coding tree, the decoder device determines the leaf node of the first-level coding tree and the node of the first-level coding tree identified by the root node of the first-level coding tree. Use the node partitioning mode corresponding to the partitioning mode information as the root node of the second-level coding tree; And parse the CTU bitstream to obtain node division mode information of the second-level coding tree.

Optionally, a method for parsing a CTU bitstream by a decoder device to obtain node split mode information of a second-level coding tree may be as follows: The decoder device parses the CTU bitstream to make second-level coding. Obtain a fourth identifier (eg, STSplitMode) used to indicate how to split the nodes of the tree, i.e., the fourth identifier represents node split mode information of the second-level coding tree.

If the fourth identifier does not indicate further splitting (eg, STSplitMode is equal to 0), then the node corresponding to the fourth identifier is a leaf node of the second-level coding tree.

If the fourth identifier indicates an additional split (eg, STSplitMode is equal to 1, 2, 3 or 4), this means that the node corresponding to the fourth identifier includes two or three child nodes, and the encoder device Further obtains a fourth identifier of the child node of the node corresponding to the fourth identifier until information about all leaf nodes of the second-level coding tree is determined. For example, if STSplitMode is 1, this indicates that the split mode corresponding to the node corresponding to the fourth identifier is horizontal binary division; If STSplitMode is 2, this indicates that the splitting mode corresponding to the node corresponding to the fourth identifier is vertical binary division; If STSplitMode is 3, this indicates that the split mode corresponding to the node corresponding to the fourth identifier is horizontal ternary split; Or STSplitMode is 4, this indicates that the split mode corresponding to the node corresponding to the fourth identifier is vertical ternary split.

Optionally, the method for parsing the CTU bitstream by the decoder apparatus to obtain node split mode information of the second-level coding tree may also be as follows: For the first node of the second-level coding tree, The decoder apparatus first parses the bitstream and then determines, according to the first preset partitioning condition, the candidate-division-mode set corresponding to the first node, and then the candidate-division-mode corresponding to the first node. The bitstream is parsed based on the set to determine split mode information corresponding to the first node.

Here, the first preset splitting condition may include a first preset splitting subcondition, a second preset splitting subcondition, a third preset splitting subcondition, a fourth preset splitting subcondition, a fifth preset splitting subcondition, And at least one of a sixth preset split subcondition, a seventh preset split subcondition, an eighth preset split subcondition, and a ninth preset split subcondition.

The first preset division subcondition is as follows: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to the first preset threshold (eg, 8 or 4), the candidate corresponding to the first node. The split-mode set does not include horizontal binary division or horizontal ternary division. The width-to-height ratio of the image area corresponding to the first node is the ratio of the width of the image area corresponding to the first node to the height of the image area corresponding to the first node.

The second preset division sub-condition is as follows: if the height-to-width ratio of the picture area corresponding to the first node is greater than or equal to the second preset threshold (eg, 8 or 4), the candidate-corresponding to the first node. The split mode set does not include vertical binary division or vertical ternary division. The height-to-width ratio of the image area corresponding to the first node is the ratio of the height of the image area corresponding to the first node to the width of the image area corresponding to the first node.

The third preset division subcondition is as follows: the ratio of the area of the image area corresponding to the first node to the area of the image area corresponding to the leaf node of the first-level coding tree to which the first node belongs is the third preset If less than or equal to a set threshold (eg, 16 or 8), the candidate-division-mode set corresponding to the first node includes horizontal binary division, horizontal ternary division, vertical binary division, or vertical ternary division. I never do that.

The fourth preset partitioning subcondition is as follows: the partitioning mode corresponding to the first node includes the first partitioning mode, and decoding of the first child node of the first node is performed by the second child node of the first child node. If it is later than decoding, the splitting mode corresponding to the second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode. The first division mode is horizontal binary division or vertical binary division.

For example, if the split mode corresponding to the first node and the second child node are both horizontal ternary splits, then the candidate-mode set corresponding to the first child node does not include the horizontal ternary split, i.e., the horizontal ternary. Splitting a first child node in split mode is not allowed.

The fifth preset division sub-condition is as follows: the division mode corresponding to the first node includes the second division mode, and the area of the image area corresponding to the first child node of the first node is 3 of the first node. If the smallest of the areas of the image area corresponding to the child nodes, the candidate-division-mode set corresponding to the first child node does not include the second division mode. The second division mode is horizontal ternary division or vertical ternary division.

For example, if the splitting mode corresponding to the first node is a horizontal ternary split, the candidate-mode set corresponding to the first child node does not include the horizontal ternary split, that is, the first in the horizontal ternary split mode. Splitting child nodes is not allowed.

The sixth preset division sub-condition is as follows: the division mode corresponding to the first node includes the second division mode, and the area of the image area corresponding to the first child node of the first node is 3 of the first node. If the largest of the areas of the image area corresponding to the child nodes, the candidate-division-mode set corresponding to the first child node does not include the second division mode.

The seventh preset division sub-condition is as follows: if the ratio of the width of the picture area corresponding to the first node to the preset side length of the minimum CU is less than or equal to the third preset threshold, the candidate corresponding to the first node; The split-mode set does not include vertical ternary splits; Or if the ratio of the height of the picture area corresponding to the first node to the preset side length of the minimum CU is less than or equal to the third preset threshold, the candidate-division-mode set corresponding to the first node includes horizontal ternary partitioning. do.

The eighth preset partitioning subcondition is as follows: if the area of the picture area corresponding to the first node is equal to or less than the fourth preset threshold, the candidate-division-mode set corresponding to the first node is horizontal binary division, horizontal; It does not include ternary splits, vertical binary splits or vertical ternary splits.

The ninth preset divisional subcondition is as follows: the division mode corresponding to the first node is the second division mode, and the area of the image area corresponding to the first child node of the first node is three children of the first node. If the largest of the areas of the image area corresponding to the node is the largest, the candidate-division-mode set corresponding to the first child node does not include the first division mode. The division direction of the first division mode is the same as the division direction of the second division mode.

For example, the division mode corresponding to the first node is horizontal ternary division, and the area of the image area corresponding to the first child node of the first node is the area of the image area corresponding to the three child nodes of the first node. In the largest case, the candidate-mode set corresponding to the first child node does not include horizontal binary division, that is, splitting the first child node in the horizontal binary division mode is not allowed.

For example, in this embodiment of the present application, the first preset division condition may include a third preset division sub condition; May include a fourth preset split subcondition, or may include a fifth preset split subcondition; May include a first preset split subcondition and a second preset split subcondition; May include a first preset split subcondition, a second preset split subcondition, a fourth preset split subcondition, a sixth preset split subcondition, and a seventh preset split subcondition; May include a first preset split subcondition, a second preset split subcondition, a third preset split subcondition, and a seventh preset split subcondition; May include a first preset split subcondition, a second preset split subcondition, a third preset split subcondition, a fifth preset split subcondition, and a seventh preset split subcondition; May include a fifth preset split subcondition and a seventh preset split subcondition; May include a third preset split subcondition, a fourth preset split subcondition, and a seventh preset split subcondition; It may also include a first preset split subcondition, a second preset split subcondition, a fourth preset split subcondition, and a fifth preset split subcondition.

Certainly, in this embodiment of the present application, the first preset partitioning condition may alternatively be another combination of the aforementioned preset partitioning subconditions. Details are not described here again.

Optionally, the encoder device is configured to determine a first preset threshold, a second preset threshold, a third preset threshold, a fourth preset threshold, and a minimum CU from the SPS, PPS, or slice header in the bitstream containing the picture data. A preset side length can be obtained. Certainly, the first preset threshold, the second preset threshold, the third preset threshold, the fourth preset threshold, and the preset side length of the minimum CU may alternatively be preset by the image processing system.

Further, the first preset division condition in this embodiment of the present application may include other conditions in addition to the first preset division sub condition to the ninth preset division sub condition. This is not particularly limited in this embodiment of the present application.

For example, the first preset splitting condition further includes any of the following conditions:

A. If the depth of the first node in the second-level coding tree is equal to the maximum depth preset in the second-level coding tree (eg, the maximum depth preset in the second-level coding tree is 3, 4, Or 2), the candidate-division-mode set corresponding to the first node includes no further partitioning.

The encoder apparatus may obtain a preset maximum second-level coding tree level from the SPS, PPS, or slice header in the bitstream including the picture data.

B. If the width of the picture area corresponding to the first node is equal to the width of the preset minimum CU, the candidate-division-mode set corresponding to the first node does not include horizontal binary division.

C. If the width or height of the picture area corresponding to the first node is greater than the fifth preset threshold (eg, 64 or 128), the candidate-division-mode set corresponding to the first node may cause non-additional division. Include.

D. When the ratio of the width of the picture area corresponding to the first node to the preset side length of the minimum CU is smaller than the sixth preset threshold, the candidate-division-mode set corresponding to the first node includes vertical ternary partitioning. Do not; Or if the ratio of the height of the picture area corresponding to the first node to the preset side length of the minimum CU is smaller than the seventh preset threshold, the candidate-division-mode set corresponding to the first node does not include horizontal ternary partitioning. Do not.

From the foregoing description, it can be seen that the fourth identifier in this embodiment of the present application is used to indicate a method of partitioning nodes of the second-level coding tree. Optionally, the fourth identifier may include first information used to indicate whether to split the node further, second information used to indicate the direction in which the node is split, and third information used to indicate the mode in which the node is split. Can be.

For example, the first information is represented by STSplitFlag. A numeric value of STSplitFlag indicates that a node corresponding to STSplitFlag is no longer split; Alternatively, if the numerical value of the STSplitFlag is equal to 1, it indicates that the node corresponding to the STSplitFlag is further split. The second information is represented by STSplitDir. If the numerical value of STSplitDir is equal to 0, it indicates that the node corresponding to STSplitDir is split in the horizontal direction; Alternatively, if the numeric value of STSplitDir is 1, it indicates that the node corresponding to STSplitDir is split vertically. The third information is represented by STSplitType. If the numeric value of STSplitType is 0, it indicates that the split mode corresponding to the node corresponding to STSplitType is binary tree split. A numeric value of STSplitDir of 1 indicates that the split mode corresponding to the node corresponding to STSplitType is a ternary tree split.

Optionally, in this embodiment of the present application, the fourth identifier may alternatively be: a binary tree splitting identifier used to indicate whether to split a node corresponding to the fourth identifier in binary tree splitting mode; In a ternary tree splitting mode, a ternary tree splitting identifier used to indicate whether to split a node corresponding to the fourth identifier; And second information (the division direction identifier is the same as the division direction identifier in the above-described example).

For example, the binary tree split identifier is represented by BTFlag. If the numeric value of the BTFlag is 0, it indicates that the partitioning mode corresponding to the node corresponding to the fourth identifier is not a BT partitioning; Or, if the numeric value of the BTFlag is 1, this indicates that the split mode corresponding to the node corresponding to the fourth identifier is BT split. The ternary tree partition identifier is TTFlag. If the numeric value of TTFlag is 0, it indicates that the partitioning mode corresponding to the node corresponding to the fourth identifier is not TT partitioning; Alternatively, if the numeric value of the TTFlag is 1, this indicates that the partition mode corresponding to the node corresponding to the fourth identifier is TT partition.

Certainly, in this embodiment of the present application, the fourth identifier may alternatively include any combination of the first information, the second information, and the third information. This is not particularly limited in this embodiment of the present application.

It is readily understood that if the splitting mode corresponding to the second-level coding tree further includes splitting modes other than binary tree splitting and ternary tree splitting, then the fourth identifier may include more information.

The fourth identifier may include first information, second information, and third information. Thus, after determining the candidate-division-mode set corresponding to the first node, the decoder device determines the first information, the second information, and the third information based on the candidate-division-mode set corresponding to the first node. . In this case, split mode information corresponding to the first node may be determined.

Specifically, the decoder device first determines a candidate-division-mode set of the first node according to the first preset division condition. The decoder apparatus then determines whether the first information, the second information, and the third information can be parsed from the bitstream based on the candidate-division-mode set of the first node. If any of the first information, second information and third information cannot be parsed from the bitstream, the numeric value of the information is a system default value.

For example, the first information is represented by STSplitFlag, the second information is represented by STSplitDir, and the third information is represented by STSplitType. If the candidate-division-mode set corresponding to the first node does not include horizontal binary division, horizontal ternary division, vertical binary division, or vertical ternary division, the division mode corresponding to the first node is no longer divided. That is, it is not necessary to parse STSplitFlag, STSplitDir and STSplitType in the bitstream, and the numeric values of STSplitFlag, STSplitDir and STSplitType are all zero. If the candidate-division-mode setter corresponding to the first node includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division, the decoder device first parses the STSplitFlag from the bitstream. . If the numeric value of STSplitFlag is 0, the numeric values of STSplitDir and STSplitType are 0 by default; Or if the numeric value of the STSplitFlag is 1, the STSplitDir is further parsed (the parsing mode of the STSplitDir is associated with the candidate-split-mode set corresponding to the first node), and the STSplitType is parsed based on the numeric value of the STSplitDir (STSplitType The parsing mode of is associated with the candidate-division-mode set corresponding to the first node). If the candidate-division-mode set corresponding to the first node includes horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division, the decoder device parses the STSplitDir and STSplitType sequentially from the bitstream, or ; Or when the candidate-division-mode set corresponding to the first node includes 1 to 3 splitting modes, at least one of STSplitDir and STSplitType may be obtained directly instead of being parsed from the bitstream.

In one example, Table 2 shows all possible cases of the candidate-division-mode set corresponding to the first node in this embodiment, and how to parse the first information, the second information and the third information corresponding to each case. Illustrated. In Table 2, if the numeric value of the Vertical Binary Tree (VBT) is 0, this indicates that the candidate-division-mode set corresponding to the first node does not include the vertical binary division; If the numeric value of the VBT is 1, this indicates that the candidate-division-mode set corresponding to the first node includes vertical binary division; If the numeric value of the Horizontal Binary Tree (HBT) is 0, it indicates that the candidate-division-mode set corresponding to the first node does not include the horizontal binary division; If the numeric value of the HBT is 1, this indicates that the candidate-division-mode set corresponding to the first node includes horizontal binary division; If the numeric value of the Vertical Ternary Tree (VTT) is 0, it indicates that the candidate-division-mode set corresponding to the first node does not include the vertical ternary partition; If the value of VTT is 1, it indicates that the candidate-division-mode set corresponding to the first node includes vertical ternary division; If the numeric value of the horizontal binary tree (HTT) is 0, it indicates that the candidate-division-mode set corresponding to the first node does not include the horizontal ternary partition; Or if the numeric value of the HTT is 1, it indicates that the candidate-division-mode set corresponding to the first node includes horizontal ternary division. a, b and c indicate that STSplitFlag, STSplitDir and STSplitType are parsed from the bitstream, respectively; NA (x) indicates setting the corresponding bit to the default value x. If parsing finds that STSplitFlag is 0, then the numeric values of STSplitDir and STSplitType are 0 by default. When parsing finds that STSplitFlag is 1, it first parses STSplitDir and then parses STSplitType based on the numeric value STSplitDir. The symbol! Is a logical NOT operation, and the expression "exp? M: n" indicates that m is used if exp is true, otherwise n is used. For example, STSplitDir? c: NA (1) indicates that if the numeric value of STSplitDir is not 0 (that is, the numeric value of STSplitDir is equal to 1. This is because in this embodiment the numeric value of STSplitDir is 0 or 1), the STSplitType is a bitstream. Obtained from; Otherwise, the numeric value of STSplitType is 1 by default. In another example,! STSplitDir? NA (0): NA (1) indicates that if the numeric value of STSplitDir is equal to 0, the numeric value of STSplitType is 0 by default, otherwise the numeric value of STSplitType is 1 by default.

Referring to Table 2, it can be seen that the decoder device can determine the STSplitFlag, STSplitDir and STSplitType based on the candidate-split-mode set corresponding to the first node.

In one example, for candidate-division-mode set 15 of Table 2, the candidate-division-mode set corresponding to the first node includes vertical binary division, horizontal binary division, horizontal ternary division, or vertical ternary division. Otherwise, the numeric values of STSplitFlag, STSplitDir, and STSplitType are all zeros by default.

In another example, for candidate-division-mode set 12 of Table 2, if the candidate-division-mode set corresponding to the first node does not include horizontal binary division, horizontal ternary division, or vertical ternary division, then decoder The device first parses the STSplitFlag from the bitstream; If the numeric value of STSplitFlag is 0, the numeric value of STSplitDir and STSplitType is 0 by default, or if the numeric value of STSplitFlag is 1, the numeric value of STSplitDir is 1 by default and the numeric value of STSplitType is 0 by default.

In another example, for candidate-division-mode set 6 of Table 2, if the candidate-division-mode set corresponding to the first node does not include horizontal ternary division or vertical ternary division, the decoder device first performs a bitstream. Parse the STSplitFlag in the; If the numeric value of STSplitFlag is 0, the numeric value of STSplitDir and STSplitType is 0 by default. If the numeric value of STSplitFlag is 1, the decoder device parses the STSplitDir in the bitstream and sets the numeric value of STSplitType to 0 by default.

Candidate-Split-Mode Set Split mode Parsing mode HBT VBT HTT VTT STSplitFlag STSplitDir
(STSplitFlag == 1) STSplitType
(STSplitFlag == 1) One One One One One a b c 2 0 One One One a b STSplitDir? c: NA (1) 3 One 0 One One a b ! STSplitDir? c: NA (1) 4 One One 0 One a b STSplitDir? c: NA (0) 5 One One One 0 a b ! STSplitDir? c: NA (0) 6 One One 0 0 a b NA (0) 7 One 0 One 0 a NA (0) c 8 0 One 0 One a NA (1) c 9 One 0 0 One a b ! STSplitDir? NA (0): NA (1) 10 0 One One 0 a b STSplitDir? NA (0): NA (1) 11 One 0 0 0 a NA (0) NA (0) 12 0 One 0 0 a NA (1) NA (0) 13 0 0 One 0 a NA (0) NA (1) 14 0 0 0 One a NA (1) NA (1) 15 0 0 0 0 NA (0) NA (0) NA (0)

In another example, for candidate-division-mode set 5 of Table 2, the first corresponding candidate-division-mode set corresponding to the first node does not include vertical ternary partitioning, and the decoder device first determines the STSplitFlag from the bitstream. Parse If the numeric value of STSplitFlag is 0, the numeric values of STSplitDir and STSplitType are 0 by default, or if the numeric value of STSplitFlag is 1, the decoder device parses the STSplitDir from the bitstream. If the numeric value of STSplitDir is 1, the numeric value of STSplitType is 0 by default, or if the numeric value of STSplitDir is 0, the decoder device parses the STSplitType from the bitstream.

After determining the numeric values of STSplitFlag, STSplitDir and STSplitType, the decoder device may determine the fourth identifier STSplitMode based on the three numeric values, that is, determine the split mode information of the first node of the second-level coding tree. have.

For example, the decoder device determines the numerical value of the STSplitMode based on the numerical values of STSplitFlag, STSplitDir and STSplitType in the mapping mode shown in Table 3.

STSplitMode STSplitFlag STSplitDir STSplitType 0 0 0 0 One One 0 0 2 One One 0 3 One 0 One 4 One One One

In one example, referring to the above example, if the numerical values of STSplitFlag, STSplitDir, and STSplitType corresponding to the first node are 1, 1, and 0, respectively, the value of STSplitMode is 2, that is, the split mode corresponding to the first node is vertical. It is a binary division.

Optionally, the candidate-division-mode set may alternatively include only horizontal ternary partitions and vertical ternary partitions, and does not include horizontal binary partitions or vertical binary partitions. In this case, the decoder device first parses the STSplitFlag from the bitstream. If the numeric value of STSplitFlag is 0, the numeric values of STSplitDir and STSplitType are 0 by default, or if the numeric value of STSplitFlag is 1, the decoder device parses the STSplitDir in the bitstream and the numeric value of STSplitType is 1 by default.

If the splitting mode of the second-level coding tree further includes splitting modes other than horizontal binary splitting, vertical binary splitting, horizontal ternary splitting, and vertical ternary splitting, the information parsing method shown in Table 2 is extended. need. This is not particularly limited in this embodiment of the present application.

From Table 2, it can be seen that among the 15 candidate-division-mode sets, the modes for parsing the first information, the second information and the third information corresponding to some candidate-division-mode sets are the same or similar. Thus, a plurality of candidate-split-mode sets can be further combined into one category, and one parsing method is used for each category. In this case, the method of parsing the first information, the second information, and the third information is the same as the parsing method of determining the first information, the second information and the third information according to Table 2, but the determining branch is Less

For example, the decoder device determines the first information STSplitFlag in the parsing mode shown in Table 4, determines the second information STSplitDir in the parsing mode shown in Table 5, and determines the third information in the parsing mode shown in Table 6. You can determine the STSplitType. && is a logical AND operation, || Is a logical OR operation,! Is a logical NOT operation. Similar to Table 2, a, b, and c indicate that the information needs to be parsed from the bitstream, and NA (x) indicates that the value of the information is set to x, respectively.

NumType> 0 NumType == 0 STSplitFlag a NA (0)

As shown in Table 4, if the quantity of split modes included in the candidate-split-mode set (ie, the first numeric value represented by NumType) is greater than zero, the decoder device parses the STSplitFlag from the bitstream; Or if the numeric value of NumType is 0, the numeric value of STSplitFlag is 0 by default. When it is determined that the numeric value of STSplitFlag is 1, the decoder device continues parsing STSplitDir and STSplitType; Otherwise, the numeric values of STSplitDir and STSplitType are 0 by default.

Referring to Table 2, in candidate-split-mode sets 1 to 14 of Table 2, it can be seen that each candidate-split-mode set includes a plurality of split modes and corresponding STSplitFlags, respectively. The candidate-split-mode set must be parsed from the bitstream. This corresponds to the case shown in Table 4. In addition, the candidate-division-mode set 15 of Table 2 does not include a division mode, that is, the number of division modes included in the candidate-division-mode set 15 is zero. The numeric value of STSplitFlag corresponding to candidate-split-mode set 15 is NA (0) by default, which corresponds to Table 4.

NumType
> = 3 NumType == 2 NumType
== 1 NumType
== 0 (HBT && VBT) || (HBT && VTT) || (VBT && HTT) (HBT && HTT) || (VBT && VTT) STSplitDir b b NA (! HBT) NA (VBT || VTT) NA (0)

As shown in Table 5, when the numeric value of NumType is greater than or equal to 3, the numeric value of NumType is equal to 2, the numeric value of NumType is equal to 1, the numeric value of NumType is equal to 0, and the decoder device Determines the STSplitDir in different modes. In addition, if the numeric value of NumType is 2, the decoder device must further determine a specific mode for determining the STSplitDir according to other conditions.

For example, if the numeric value of NumType is greater than 3, the decoder device analyzes the bitstream to obtain STSplitDir. If the numeric value of NumType is 2 and the condition (HBT && HTT) || (VBT && VTT) is satisfied, then the numeric value of STSplitDir is set to! HBT (that is, if HBT is 0, STSplitDir is 1, or HBT is 1 If STSplitDir is 0).

Referring to Table 2, in the candidate-split-mode sets 1 to 5 of Table 2, the quantity of split modes included in each candidate-split-mode set is greater than or equal to 3, and each candidate-split-mode set It can be seen that the STSplitDir corresponding to 1 to 5 need to be parsed from the bitstream. This corresponds to the case shown in Table 5. In the candidate-division-mode sets 6 to 10 of Table 2, the quantity of division modes included in each candidate-division-mode set is equal to 2, and under different conditions, corresponding to all candidate-division-mode sets 6 to 10 There are different modes for parsing STSplitDir, which corresponds to the case shown in Table 5. In candidate-division-mode sets 11 to 14 of Table 2, the number of split modes included in each candidate-division-mode set is equal to 1, and the numeric value of STSplitDir corresponding to each of the candidate-division-mode sets 11 to 14 respectively. Is the default value NA (VBT || VTT), which corresponds to the case shown in Table 5. The number of split modes included in candidate split mode-mode set 15 of Table 2 is equal to 0, and the STSplitDir corresponding to candidate split mode-mode set 15 is the default value NA (0), which corresponds to the case shown in Table 5. .

NumType
== 4 NumType
== 0 NumType == 3 (! HBT ||! HTT) && STSplitDir (! VBT ||! VTT) &&
! STSplitDir (! HBT ||! HTT) &&! STSplitDir (! VBT ||! VTT) &&
STSplitDir STSplitType c NA (0) c c NA (! HBT) NA (! VBT) NumType == 1 NumType == 2 (! HTT &&! VTT) (HBT && HTT) || (VBT && VTT) (HBT && VTT) (VBT && HTT) STSplitType NA (HBT || VBT) NA (0) c NA (STSplitDir) NA (! STSplitDir)

As shown in Table 6, if the numeric value of NumType is 4, the numeric value of NumType is 3, the numeric value of NumType is 2, the numeric value of NumType is 1, and the value of numeric NumType is 0, the decoder device Determines the STSplitType in another mode. Also, if the numeric value of NumType is 2 and the numeric value of NumType is 3, the decoder device must determine a specific mode for determining the STSplitType according to other conditions. Similar to Tables 4 and 5, Table 6 also corresponds to Table 2.

Since the method of determining the STSplitType by the decoder device according to Table 6 is similar to the method of determining the STSplitDir by the decoder device according to Table 5, it is not described herein again.

If the split information of the first-level coding tree is represented by NSFlag and QTSplitFlag, when the numeric value of the NSFlag of the node is the first value, the node is a leaf node of the second-level coding tree, and the numerical value of the STSplitFlag is by default. When the numeric value of the node's NSFlag is the second value and the numeric value of the QTSplitFlag is the second value, the node is not a leaf node of the second-level coding tree and the numeric value of the STSplitFlag is 1 by default.

Optionally, in this embodiment of the present application, after performing S601, the decoder device performs S602; Or in the process of performing S601, after obtaining information about leaf nodes of the first-level coding tree, and based on leaf nodes of the first-level coding tree until the last leaf node of the first-level coding tree is obtained. S602 can be performed immediately.

S603. If the node splitting mode information of the second-level coding tree indicates that the splitting mode corresponding to the first node of the second-level coding tree is non-additional splitting, the decoder device parses the bitstream to obtain encoding information of the first node. Acquire.

In this embodiment of the present application, if the node splitting mode information of the second-level coding tree indicates that the splitting mode corresponding to the first node of the second-level coding tree is no longer splitted, Indicates that this is a leaf node of a two-level coding tree.

From the foregoing description, it can be seen that a node that is no longer partitioned corresponds to one CU. Correspondingly, the first node corresponds to one CU.

Specifically, the decoder device parses the syntax structure of the coding unit in the bitstream (for example, the syntax structure coding_unit () of H.265. For the description of the syntax structure coding_unit (), see the following description). Obtains encoding information. The encoding information of each CU includes information such as prediction mode and transform coefficients of the CU. Here, that the decoder device obtains encoding information of each CU means that the decoder device obtains encoding information of each leaf node of the second-level coding tree.

Optionally, in this embodiment of the present application, in the process of performing S602, after obtaining information regarding leaf nodes of the second-level coding tree, the decoder device parses the bitstream to obtain encoding information of the leaf nodes. Can do it; After obtaining the encoding information of the leaf node, encoding information of the leaf node of the next second-level coding tree and the leaf node of the next second-level coding tree until the last second-level coding tree is obtained. Continue to get back.

For example, the syntax table of the second-level coding tree in this embodiment of the present application is shown in Table 7. coding_second_tree () is a syntax structure of the second-level coding tree and describes information about nodes of the second-level coding tree.

In Table 7, log2CuWidth and log2CuHeight represent log values of the image area corresponding to the node having a base of 2 and log values of the height of the image area corresponding to the node having a base of 2, respectively, and stDepth represents the log value of the second-level coding tree. Depth of the leaf node, condition condB indicates the condition that the syntax element STSplitMode of the node split information of the second-level coding tree needs to be parsed from the bitstream. For example, the condition condB is " the depth stDepth of the node in the second-level coding tree is less than the preset maximum depth maxSTDepth in the second-level coding tree, and both the width or height of the picture area corresponding to the node is less than the threshold maxSTSize. Small and the width or height of the image area corresponding to the node is greater than the threshold minSTSize ". maxSTDepth is an integer greater than zero (eg, 2, 3 or 4) and can be preset or parsed in the SPS; minSTSize is an integer greater than zero (eg, 4 or 8) and can be preset or parsed in the SPS; maxSTSize is an integer greater than minSTSize (eg, 64 or 128) and can be preset or parsed in the SPS.

The range of values for STSplitMode is 0, 1, 2, 3, and 4. If the numeric value of STSplitMode is 0, the node is a leaf node of the second-level coding tree and the node corresponds to one CU. In this case, the information about the coding unit is parsed based on the CU syntax structure coding_unit (). The mode of constructing the syntax element of the information about the coding unit is not limited in this embodiment of the present application. If the numeric value of STSplitMode is 1-4, the node is split into two or three child nodes individually through horizontal binary division, vertical binary division, horizontal ternary division, and vertical ternary division. The width, height and coordinates of the image region corresponding to the child node, and the depth in the second-level coding tree are determined, and the child nodes are parsed sequentially based on coding_second_tree ().

Optionally, in this embodiment of the present application, the STSplitFlag, STSplitDir and STSplitType may be encoded in a by-pass manner, or may be encoded using one probability model, or the probability model may be based on a context. It can be chosen adaptively.

For example, the method of selecting the context model of the STSplitFlag is as follows: the size S1 of the CU corresponding to the left side of the picture area corresponding to the node, the size S2 of the CU corresponding to the upper side of the picture area corresponding to the node and The size S3 of the picture area corresponding to the node is obtained. If both S1 and S2 are less than S3, the numbered context model is selected; Or if only one of S1 and S2 is smaller than S3, the numbered context model number 1 is selected; Otherwise (if neither S1 nor S2 is less than S3) the numbered context model is selected. The left side of the image region corresponding to the node is (x0-1, y0), for example, and the upper side of the image region corresponding to the node is (x0, y0-1), for example.

The method for selecting the context model of STSplitDir is as follows: If the width of the picture area corresponding to the node is twice as large as the height, the numbered model is used; If the width of the picture area corresponding to the node is twice the height, the numbered model 1 is used; If the width of the picture area corresponding to the node is equal to the height, the numbered model 2 is used; If the width of the picture area corresponding to the node is equal to 1/2 of the height, the numbered model 3 is used; Or if the width of the image area corresponding to the node is smaller than 1/2 of the height, the model numbered 4 is used.

The method of selecting the context model of the STSplitType is as follows: if the depth of the node is zero in the second-level coding tree, the numbered model is selected; If the depth of the node in the second-level coding tree is 1, the numbered model is selected; Or if the depth of the node in the second-level coding tree is greater than one, the numbered model 2 is selected.

It should be understood that the number of context models described above is only used to distinguish different context models, and the number of context models is not limited to the same number as in the above example.

coding_second_tree (x0, y0, log2CuWidth, log2CuHeight, stDepth, qtDepth) { Descriptor if (condB) STSplitMode [x0] [y0] ae (v)   if (STSplitMode [x0] [y0] == 1) {     y1 = y0 + (1 << (log2CuHeight-1))    coding_second_tree (x0, y0, log2CuWidth, log2CuHeight1, stDepth + 1, qtDepth)    coding_second_tree (x0, y1, log2CuWidth, log2CuHeight1, stDepth + 1, qtDepth)   }   else if (STSplitMode [x0] [y0] == 2) {     x1 = x0 + (1 << (log2CuWidth-1))      coding_second_tree (x0, y0, log2CuWidth1, log2CuHeight, stDepth + 1, qtDepth)    coding_second_tree (x1, y0, log2CuWidth1, log2CuHeight, stDepth + 1, qtDepth)   }   else if (STSplitMode [x0] [y0] == 3) {     y1 = y0 + (1 << (log2CuHeight-2))     y2 = y1 + (1 << (log2CuHeight-1))  coding_unit (x0, y0, log2CuWidth, log2CuHeight2, stDepth + 1, qtDepth)  coding_unit (x0, y0, log2CuWidth, log2CuHeight1, stDepth + 1, qtDepth)  coding_unit (x0, y2, log2CuWidth, log2CuHeight2, stDepth + 1, qtDepth)   }   else if (STSplitMode [x0] [y0] == 4) {     x1 = x0 + (1 << (log2CuWidth-2))     x2 = x1 + (1 << (log2CuWidth-1)) coding_unit (x0, y0, log2CuWidth 2, log2CuHeight, stDepth + 1, qtDepth) coding_unit (x1, y0, log2CuWidth 1, log2CuHeight, stDepth + 1, qtDepth) coding_unit (x2, y0, log2CuWidth 2, log2CuHeight, stDepth + 1, qtDepth)   }   else     coding_unit (x0, y0, log2CuWidth, log2CuHeight, stDepth, qtDepth) }

S604. The decoder device decodes and reconstructs the coding unit based on the encoding information of the three child nodes of the first node to obtain an image corresponding to the picture data.

The process of decoding and reconstructing the coding unit includes processing such as prediction, inverse quantization, inverse transform and loop filtering. Specifically, the process of decoding and reconstructing each coding unit includes the following steps:

(1) Intra prediction or inter prediction is selected based on the prediction mode included in the encoding information of the CU to obtain the predicted pixels of the CU.

(2) if the CU includes transform coefficients, inverse quantization and inverse transform are performed on the transform coefficients of the CU based on the quantization parameter and the transform mode to obtain a reconstructed residual of the CU; Or if the CU does not include transform coefficients, the reconstructed residual of the CU is zero, that is, the reconstructed residual value of each pixel in the CU is zero.

(3) After the predicted pixel of the CU and the reconstructed residual of the CU are added, a loop filtering process is performed to obtain a reconstructed picture block of the CU.

The encoder apparatus decodes and reconstructs each coding unit according to the method described above to obtain a reconstructed picture block of each CU. After obtaining the reconstructed picture block of each CU, the encoder device obtains the final reconstructed picture based on all the reconstructed picture blocks obtained, that is, obtains a picture corresponding to the picture data.

In this embodiment of the present application, the splitting mode corresponding to the first node of the second-level coding tree is one of splitting modes determined according to the first preset splitting condition, and the first preset splitting condition is determined by the first node. Split in the target division mode, the target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. The first preset partitioning condition reduces the quantity of partitioning modes corresponding to the first node, effectively reducing the decoding complexity.

In the embodiment shown in FIG. 6, in the process of determining node split mode information of the second-level coding tree, the decoder device may sequentially determine the first information, the second information, and the third information. The decoder device in this embodiment of the present application may further determine node division mode information of the second-level coding tree in the sequence of the first information, the third information, and the second information. This scenario is now described.

For example, in S602, the decoder device further sequentially determines the first information, the third information and the second information according to the parsing method shown in Table 8, so that the second level based on the mapping mode shown in Table 3 is obtained. Node splitting mode information of the coding tree may be obtained.

Table 8 is similar to Table 2. The detailed description is not described herein again.

Similar to Table 2, among the 15 candidate-division-mode sets shown in Table 8, the modes for parsing the first information, the third information, and the second information corresponding to some candidate-division-mode sets are the same or Or similar. Thus, some of the candidate-division-mode sets can be further combined into one category, and one parsing method is used per category. In this case, the method of parsing the first information, the third information and the second information is the same as the parsing method for determining the first information, the third information and the second information according to Table 8, but with fewer decision branches.

For example, the decoder device determines the first information STSplitFlag based on the parsing mode shown in Table 4, determines the third information STSplitType based on the parsing mode shown in Table 9, and determines the parsing shown in Table 10. Based on the second information STSplitDir can be determined.

Candidate-Split-Mode Set Split mode Parsing mode HBT VBT HTT VTT STSplitFlag STSplitType (STSplitFlag == 1) STSplitDir
(STSplitFlag == 1) One One One One One a c b 2 0 One One One a c STSplitType? b: NA (1) 3 One 0 One One a c STSplitType? b: NA (0) 4 One One 0 One a c ! STSplitType? b: NA (1) 5 One One One 0 a c ! STSplitType? b: NA (0) 6 One One 0 0 a NA (0) b 7 One 0 One 0 a c NA (0) 8 0 One 0 One a c NA (1) 9 One 0 0 One a c ! STSplitType? NA (0): NA (1) 10 0 One One 0 a c STSplitType? NA (0): NA (1) 11 One 0 0 0 a NA (0) NA (0) 12 0 One 0 0 a NA (0) NA (1) 13 0 0 One 0 a NA (1) NA (0) 14 0 0 0 One a NA (1) NA (1) 15 0 0 0 0 NA (0) NA (0) NA (0)

NumType
> = 3 NumType == 2 NumType
== 1 NumType == 0 (HBT && VBT) (HBT && HTT)
(VBT && VTT)
(HBT && VTT)
(VBT && HTT) STSplitType c NA (0) c NA (HBT || VBT) NA (0)

NumType == 4 NumType == 0 NumType == 3 (! HBT ||! VBT) && STSplitType (! HTT ||! VTT) &&
! STSplitType (! HBT ||! VBT) &&
! STSplitType (! HTT ||! VTT) && STSplitType STSplitDir b NA
(0) b b NA (VBT) NA (VTT) NumType == 1 NumType == 2 (! HTT &&! VTT) (HBT && HTT) || (VBT && VTT) (HBT && VTT) (VBT && HTT) STSplitDir NA (HTT || VTT) b NA (VBT) NA (STSplitType) NA (! STSplitType)

As can be seen, the decoder device determines the node split mode information of the second-level coding tree in the sequence of the first information, the second information and the third information, or the first information, the third information and the second information. Regardless of whether the node split mode information of the second-level coding tree is determined in the sequence of, the decoder device may perform fast decoding as a function of the first preset splitting condition and the decoding complexity is relatively low.

In the above embodiment, the splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary tree splitting. . In a practical application, the splitting mode corresponding to the second-level coding tree may further comprise quadtree splitting. In this case, the decoder apparatus should determine the candidate-division-mode set corresponding to the first node according to the first preset division condition and the second preset division condition, and determine the division mode corresponding to the first node. .

As shown in Fig. 7, in this scenario, the image data decoding method provided in this embodiment of the present application includes the following steps.

S700. The decoder device obtains a bitstream containing the image data.

For S700, refer to the above description of S600. Details are not described here.

S701. The decoder device decodes the bitstream obtained by the decoder device to obtain node division mode information of the first-level coding tree.

For S701, the description of S601 described above is referred to. Details are not described here again.

S702. The decoder device parses the bitstream obtained by the decoder device to obtain node division mode information of the second-level coding tree.

Node splitting mode information of the second-level coding tree indicates a splitting mode corresponding to the first node of the second-level coding tree, and a splitting mode corresponding to the first node is a candidate-split-mode corresponding to the first node. One mode in the set, and the candidate-division-mode set corresponding to the first node is determined according to the first preset splitting condition and the second preset splitting condition, wherein the first preset splitting condition is determined by the first node. Used to indicate whether to restrict splitting in split mode, the target split mode may include at least one of a horizontal binary split, a horizontal ternary split, a vertical binary split, and a vertical ternary split. And a second preset splitting condition is used to indicate whether the first node is to be restricted from splitting based on quadtree splitting.

The root node of the second-level coding tree is a leaf node of the first-level coding tree. The node splitting mode corresponding to the second-level coding tree is different from the node splitting mode corresponding to the first-level coding tree. In this embodiment, the node splitting mode corresponding to the second-level coding tree includes binary tree splitting, ternary tree splitting, and quadtree splitting, wherein the binary tree splitting is horizontal binary splitting and vertical binary splitting. And a ternary tree split includes a horizontal ternary split and a vertical ternary split.

Unlike S602, the decoder apparatus of this embodiment must determine the candidate-division-mode set corresponding to the first node according to the first preset division condition and the second preset division condition.

The first preset dividing condition is the same as the first preset dividing condition of the embodiment shown in FIG. The second preset split condition is a preset split subcondition:

10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the candidate-division corresponding to the first node. A condition set does not include the quadtree splitting; And

Eleventh preset partitioning subcondition: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is divided into the quadtrees. Conditions that do not include

At least one of the.

Similar to that of S602, the decoder apparatus of the present embodiment first determines, according to the first preset division condition and the second preset division condition, the candidate-division-mode set corresponding to the first node, The bitstream is parsed in the corresponding partitioning mode to determine node partitioning mode information corresponding to the second-level coding tree.

The node splitting mode corresponding to the second-level coding tree in this embodiment further includes quadtree splitting in this embodiment, based on the description in FIG. 6, and the numeric value of the fourth identifier STSplitMode is also 5 May be STSplitMode = 5, which indicates that the split mode corresponding to the node corresponding to the fourth identifier is quadtree split.

Correspondingly, the fourth identifier of the present embodiment further includes fourth information (in this specification, the fourth information is represented by STQTSplitFlag), and the fourth information indicates whether to split the first node based on quadtree splitting. Used. In other words, the fourth identifier of the present embodiment includes first information, second information, third information, and fourth information. In this embodiment, if the numeric value of STQTSplitFlag is 1, it indicates that the splitting mode corresponding to the node corresponding to the fourth identifier includes quadtree splitting; Or if the numeric value of STQTSplitFlag is 0, this indicates that the candidate-split-mode set corresponding to the node corresponding to the fourth identifier does not include quadtree splitting.

In detail, the decoder device sequentially determines the first information, the second information, the third information, and the fourth information to determine the node division mode information corresponding to the second-level coding tree.

Similarly to that in the embodiment of FIG. 6, the decoder device of the present embodiment determines the first information based on the content of the first numeric value and the fourth information in the embodiment of FIG. The third information can be determined.

For example, the decoder device may determine the first information STSplitFlag in the parsing mode shown in Table 11, and determine the fourth information STQTSplitFlag in the parsing mode shown in Table 12. In Table 11, if the numerical value of QT is 1, it indicates that the first node is split in the QT splitting mode; Or if the numerical value of QT is 0, it indicates that the first node split is not allowed in the QT split mode. If "NumType> 0 QT" is satisfied, the decoder device parses the STSplitFlag from the bitstream; Otherwise (ie, if "NumType == 0 &&! QT" is satisfied), the numeric value of STSplitFlag is 0 by default. If the numeric value of STSplitFlag is 1, the decoder device continues parsing STQTSplitFlag; Otherwise, the numeric values of STQTSplitFlag, STSplitDir, and STSplitType are all zeros by default.

From Table 12, when the splitting mode corresponding to the first node includes quadtree splitting, the decoder device parses the STQTSplitFlag from the bitstream; Otherwise, it can be seen that the numeric value of STQTSplitFlag is 0 by default. If the numeric value of STQTSplitFlag is 0, the STSplitDir and STSplitType continue to be parsed sequentially, otherwise the numeric value of STSplitDir and STSplitType is 0 by default.

In this embodiment, the mode of parsing STSplitDir and STSplitType is the same as the mode of parsing STSplitDir and STSplitType in the embodiment of FIG. Details are not described here again.

NumType> 0 || QT NumType == 0 &&! QT STSplitFlag a NA (0)

QT ! QT STQTSplitFlag a NA (0)

After determining the numerical values of STSplitFlag, STSplitDir, STSplitType, and STQTSplitFlag, the decoder device may determine the fourth identifier STSplitMode based on the four numerical values, that is, the split mode of the first node of the second-level coding tree. Information can be determined.

For example, the decoder device determines the numerical value of STSplitMode based on the numerical values of STSplitFlag, STSplitDir, STSplitType, and STQTSplitFlag in the mapping mode shown in Table 13.

STSplitMode STSplitFlag STQTSplitFlag STSplitDir STSplitType 0 0 0 0 0 One One 0 0 0 2 One 0 One 0 3 One 0 0 One 4 One 0 One One 5 One One 0 0

As with other information, the STQTSplitFlag in this embodiment may be encoded in bypass mode or encoded using one or more probability models.

S703. If the node splitting mode information of the second-level coding tree indicates that the splitting mode corresponding to the first node of the second-level coding tree is non-additional splitting, the decoder device parses the bitstream to obtain encoding information of the first node. Acquire.

See S603 for S703. Details are not described here again.

The decoder device decodes and reconstructs the coding unit based on the encoding information of the three child nodes of the first node to obtain an image corresponding to the picture data.

See S604 for S704. Details are not described here again.

Compared with the embodiment of FIG. 6, in this embodiment, the splitting mode corresponding to the second-level coding tree further includes quadtree splitting, and the splitting included in the candidate-split-mode set corresponding to the first node. The mode is more limited. In this way, the decoding efficiency of the decoder device can be further improved.

In addition, embodiments of the present application further provide a method of encoding image data. The encoding method can be applied to the image processing system shown in FIG.

As shown in Fig. 8, the image data encoding method provided by this embodiment of the present application includes the following steps.

S800. The encoder device determines the CTU corresponding to the picture block to be encoded.

From the foregoing description, it can be seen that an image includes a plurality of CTUs, and one CTU generally corresponds to one square image area. After obtaining the picture, the encoder device encodes each CTU of the picture.

The process of encoding all CTUs by the encoder device is the same. Thus, a description is provided in this embodiment of the present application by using an example in which the encoder device encodes one CTU.

S801. The encoder apparatus obtains a leaf node of the first-level coding tree by dividing the CTU in the node division mode corresponding to the first-level coding tree.

The root node of the first-level coding tree corresponds to the CTU, and the node splitting mode corresponding to the first-level coding tree is quadtree splitting.

Specifically, the encoder apparatus determines the CTU as the root node of the first-level coding tree, and recursively splits the CTU into at least one leaf node in the QT partitioning mode.

After obtaining at least one leaf node of the first-level coding tree, the encoder apparatus determines each leaf node of the first-level coding tree as the root node of the second-level coding tree, and the second-level coding tree The following steps are sequentially performed for each root node of the second-level coding tree until information for all leaf nodes of is obtained. Description is provided in this embodiment of the present application by using an example in which the encoder apparatus processes a first node of a second-level coding tree.

S802. The encoder apparatus determines a candidate-division-mode set corresponding to the first node of the second-level coding tree, and the candidate-division-mode set corresponding to the first node satisfies the first preset division condition.

The first preset splitting condition is that the first node is split in the target splitting mode, the root node of the second-level coding tree is a leaf node of the first-level coding tree, and the target splitting mode is horizontal binary splitting, horizontal 3 At least one of binary division, vertical binary division, and vertical ternary division.

In general, if there are no restrictions on the splitting of the first node, the splitting modes available for the first node are no further splitting, horizontal binary splitting, horizontal ternary splitting, vertical binary splitting, vertical 3 It includes binary partitioning and quadtree partitioning. In this case, the encoder device must calculate the RD costs of the six split modes sequentially, resulting in a relatively high encoding complexity.

This embodiment of the present application proposes a first preset partitioning condition for limiting the partitioning mode available to the first node. The first preset dividing condition is the same as the first preset dividing condition described in the embodiment shown in FIG. Details are not described here again.

The quantity of splitting modes included in the candidate-split-mode set corresponding to the first node is limited by the first preset splitting condition. In this way, the encoder device reduces the encoding complexity since in most cases it is not necessary to calculate the RD cost of the six split modes for the first node.

S803. If the candidate-division-mode set corresponding to the first node includes non-additional division, the encoder device encodes the CU corresponding to the first node to obtain a coding unit bitstream corresponding to the CU.

If the candidate-division-mode set corresponding to the first node includes non-additional splitting, this indicates that the first node is a leaf node of the second-level coding tree and the first node corresponds to one CU. The encoder device encodes a CU corresponding to the first node to obtain a coding unit bitstream corresponding to the CU.

Specifically, CU encoding includes procedures such as prediction, transform, quantization, and entropy coding. In the case of a CU, the process by which the encoder device encodes a CU to obtain a CU bitstream corresponding to the CU mainly includes the following steps:

(1) The encoder apparatus selects intra prediction or inter prediction based on the prediction mode to obtain predicted pixels of the CU.

(2). The encoder device obtains transform coefficients by changing and quantizing the residual between the original pixel of the CU and the predicted pixel of the CU, and performs inverse quantization and inverse transform processing on the obtained transform coefficients to obtain a reconstructed residual of the CU.

(3) After adding the predicted pixels of the CU and the reconstructed residual of the CU, the encoder device performs a loop filtering process to obtain the reconstructed pixels of the CU.

(4) The encoder device performs entropy coding on information such as a prediction mode and transform coefficients of the CU to obtain a bitstream of the CU.

For the detailed process of the encoder device encoding the CU to obtain the CU bitstream of the CU, refer to the existing CU bitstream generation method, and details are not described herein again.

S804. If the candidate-division-mode set corresponding to the first node does not include non-additional division, the encoder apparatus calculates the RD cost of each division mode in the candidate-division-mode set corresponding to the first node.

For one candidate-division-mode set of one of the candidate-division-mode sets corresponding to the first node, the encoder apparatus divides the first node in this division mode, and obtains after the first node is divided in this division mode. Get all CUs. The encoder device calculates the RD cost of each CU and determines the sum of the RD costs of all CUs as the RD cost of this split mode.

Optionally, for any CU, the RD cost of the CU is estimated by the sum of the squared errors (SSE) of the reconstruction distortion of the pixels included in the CU and the quantity of bits in the bitstream corresponding to the CU. Equivalent to the weighted sum of the values.

S805. The encoder apparatus determines the division mode corresponding to the minimum rate distortion cost as the target division mode corresponding to the first node.

S806. The encoder device splits the first node in the target splitting mode corresponding to the first node.

After dividing the first node in the target splitting mode corresponding to the first node, the encoder apparatus sequentially processes S802 to S806 for each child node of the first node until all leaf nodes of the second-level coding tree are obtained. To be done.

Each leaf node of the second-level coding tree corresponds to one CU. After obtaining all leaf nodes of the second-level coding tree, the encoder apparatus can obtain a CU corresponding to each leaf node of the second-level coding tree. From S803, it can be seen that the encoder device can obtain a CU bitstream corresponding to the CU by encoding the CU. Thus, the encoder device may obtain at least one CU bitstream. In this way, the encoder apparatus obtains the CTU bitstream based on at least one CU bitstream, node split mode information corresponding to the first-level coding tree, and node split mode information corresponding to the second-level coding tree. A bitstream including image data can be generated.

In addition, the encoder apparatus in this embodiment of the present application may further determine a candidate-division-mode set corresponding to the first node according to the second preset division condition.

Here, the second preset splitting condition may include at least one of the following preset splitting subconditions.

10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than the preset maximum depth in the second-level coding tree, the candidate-division-mode set corresponding to the first node is quad. Does not include tree splitting.

Eleventh Preset Split Subcondition: If the width-to-height ratio of the picture region corresponding to the first node is greater than the fifth preset threshold, the candidate-division-mode set corresponding to the first preset does not include quadtree splitting. .

In this embodiment of the present application, the partitioning of the nodes of the second-level coding tree is limited by the first preset condition. In this way, the complexity of splitting the nodes of the second-level coding tree is greatly reduced, and the encoding complexity is reduced.

An embodiment of the present application provides a decoder device. The decoder device is configured to perform the steps performed by the decoder device in the above-described image data decoding method. The decoder device provided by this embodiment of the present application may include a module corresponding to the corresponding step.

In this embodiment of the present application, the division of the functional module of the decoder device may be performed according to the example of the above-described method. For example, partitioning of function modules may be performed in correspondence with functions, or two or more functions may be integrated into one processing module. The integrated module may be implemented in the form of hardware or in the form of a software function module. In this embodiment of the present application, the module division is merely an example, logical function division, and may be other division in actual implementation.

When division of the function module is performed corresponding to the function, FIG. 9 is a schematic structural diagram of the terminal device in the above-described embodiment. As shown in FIG. 9, the decoder apparatus includes an acquisition module 900, a parsing module 901, and a decoding and reconstruction module 910. The acquisition module 900 is configured to support the decoder device in performing S600, S700, etc. in the above-described embodiments, and / or is configured to perform other processes of the techniques described in this application. The parsing module 901 is configured to support the decoder device in performing S601, S602, S603, S701, S702, S703, etc. in the above-described embodiments, and / or performs other processes of the techniques described in this application. It is configured to. The decoding and reconstruction module 910 is configured to support the decoder device in performing S604, S704, etc. in the above-described embodiment, and / or is configured to perform other processes of the techniques described in this application. All relevant details of the steps in the foregoing method embodiments may be cited in the functional description of the corresponding functional module, and details are not described herein again. Certainly, the decoder device provided by this embodiment of the present application includes, but is not limited to, the aforementioned module. For example, the decoder device may further include a storage module 911. The storage module 911 may be configured to store program code and data of the decoder device.

When an integrated unit is used, the parsing module 901 and the decoding and reconstruction module 910 in this embodiment of the present application may be the processor 57 in FIG. 5. The acquisition module 900 may be the RF circuit 50 in FIG. 5, and the antenna and storage module 911 connected to the RF circuit 50 may be the memory 51 in FIG. 5.

When the decoder device is executed, the decoder device performs the image data decoding method in the embodiment shown in Fig. 6 or 7. For a specific picture data decoding method, refer to the related description of the above-described embodiment shown in FIG. 6 or FIG. 7. Details are not described here again.

Another embodiment of the present application further provides a computer readable storage medium. Computer-readable storage media includes one or more program code groups, and one or more programs include instructions. When the processor in the decoder device executes the program code, the decoder device performs the image data decoding method shown in FIG. 6 or 7.

Another embodiment of the present application further provides a computer program product. The computer program product includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. When at least one processor of the decoder device is capable of reading computer executable instructions from a computer readable storage medium, the at least one processor executes the computer executable instructions, so that the decoder device is capable of the image data shown in FIG. 6 or 7. Performing the steps of the decoder device in the decoding method.

An embodiment of the present application provides an encoder device. The encoder apparatus is configured to perform the steps performed by the encoder apparatus in the above-described image data decoding method. The encoder device provided by this embodiment of the present application may comprise a module corresponding to the corresponding step.

In this embodiment of the present application, the division of the functional module may be performed in the encoder apparatus according to the example of the above-described method. For example, partitioning of function modules may be performed in correspondence with functions, or two or more functions may be integrated into one processing module. The integrated module may be implemented in the form of hardware or in the form of a software function module. In this embodiment of the present application, the module division is merely an example, logical function division, and may be other division in actual implementation.

When division of the function module is performed corresponding to the function, FIG. 10 is a possible schematic structural diagram of the terminal device in the above-described embodiment. As shown in FIG. 10, the encoder apparatus includes a determination module 1000, a division module 1001, an encoding module 1010, and a calculation module 1011. The determination module 1000 is configured to support the encoder device in performing S800, S802, S805, etc. in the above-described embodiments and / or to perform other processes in the techniques described herein. The splitting module 1001 is configured to support the encoder device in performing S801, S806, etc. in the above-described embodiment and / or to perform other processes in the techniques described herein. The encoding module 1010 is configured to support the encoder device in performing S803 in the above-described embodiment and / or to perform other processes in the techniques described herein. The calculation module 1011 is configured to support the encoder device in performing S804 in the above-described embodiment and / or to perform other processes in the techniques described herein. All relevant details of the steps in the foregoing method embodiments may be cited in the functional description of the corresponding functional module, and details are not described herein again. Certainly, the encoder device provided by this embodiment of the present application includes, but is not limited to, the module described above. For example, the encoder device may further include a storage module 1002, a transmission module 1003, and a reception module 1004. The storage module 1002 may be configured to store program code and data of the encoder device. The transmitting module 1003 and the receiving module 1004 are configured to communicate with other devices.

When the integration unit is used in this embodiment of the present application, the determination module 1000, the division module 1001, the encoding module 1010 and the calculation module 1011 may be the processor 57 in FIG. 5. The transmitting module 1003 and the receiving module 1004 may be the RF circuit 50 in FIG. 5. The antenna and storage module 1002 coupled to the RF circuit 50 may be the memory 51 in FIG. 5.

When the encoder device is executed, the encoder device performs the image data decoding method in the embodiment shown in FIG. For a specific picture data decoding method, refer to the related description of the above-described embodiment shown in FIG. Details are not described herein again.

Another embodiment of the present application further provides a computer readable storage medium. Computer-readable storage media includes one or more program code groups, and one or more programs include instructions. When the processor of the encoder device executes the program code, the encoder device performs the image data decoding method shown in FIG.

In another embodiment of the present application, a computer program product is further provided. The computer program product includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. At least one processor of the encoder device may read computer executable instructions from a computer readable storage medium, and the at least one processor executes the computer executable instructions so that the encoder device may be used in the image data decoding method shown in FIG. Perform the steps of the encoder device.

All or part of the above-described embodiments may be implemented using software, hardware, firmware or any combination thereof. When a software program is used to implement an embodiment, the embodiment may be fully or partially implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When computer program instructions are loaded into a computer and executed, procedures or functions in accordance with embodiments of the present application are generated in whole or in part. The computer may be a general purpose computer, special purpose computer, computer network or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from the computer readable storage medium to another computer readable storage medium. For example, computer instructions may be wired (e.g., coaxial cable, optical fiber, or digital subscriber line (DLS)), or wireless (e.g., infrared, radio, or microwave) for a website, computer, server. Or from a data center to another web site, computer, server, or data center. Computer-readable storage media can be any available media that can be accessed by a computer, or a data storage device such as a server or data center incorporating one or more available media. Usable media can be magnetic media (eg, floppy disks, hard disks or magnetic tape), optical media (eg, DVD), semiconductor media (eg, solid state disks (SSD)), and the like. .

With the description of the foregoing implementations, those skilled in the art can clearly understand that the division of the above-described functional modules is used only as an example of description for convenience and brevity of description. In practical application, the above-described functions may be assigned to different function modules to implement as needed, that is, the internal structure of the apparatus is divided into different function modules to implement all or part of the above-described functions.

In the several embodiments provided in this application, it goes without saying that the system, apparatus, and method described above can be realized in other ways as well. For example, the described device embodiments are merely illustrative. For example, the division of units is merely a sort of logical function division, and may be in a different division manner during actual execution. For example, a plurality of units or components may be combined or integrated into another system, or some features may be ignored or not performed. In addition, the illustrated or discussed mutual couplings or direct couplings or communication connections may be realized through some interfaces. Indirect coupling or communication connections between devices or units may be realized in electronic, mechanical or other forms.

Units described as separate parts may or may not be physically separate, and parts shown as units may or may not be physical units, may be located in one location, or may be distributed across a plurality of network units. . Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, the functional units in the embodiment of the present invention may be integrated into one processing unit, each unit may exist physically alone, or two or more units may be integrated into one unit. The integration unit may be realized in the form of hardware or in the form of a software functional unit.

If the integrated unit is realized in the form of a software functional unit and is marketed or used as a standalone product, the integrated unit can be stored in a computer readable storage medium. Based on this understanding, the essential technical solution of the present invention or the part contributing to the prior art, or part of the technical solution can be realized in the form of a software product. The computer software product may be stored on a storage medium and instruct a computer device (which may be a personal computer, a server, or a network device, etc.) to perform some or all of the steps of the method described in embodiments of the present invention. Contains two commands. The above-mentioned storage medium may be any storage medium capable of storing program code, for example, a USB flash disk, a portable hard disk, a read only memory (ROM), a random access memory (RAM). Magnetic disks or optical disks.

The foregoing descriptions are merely specific implementation manners of the present invention, but are not intended to limit the protection scope of the present invention. All variations or replacements readily realized by those skilled in the art within the technical scope described in the present invention shall fall within the protection scope of the present invention. Therefore, the protection scope of the present invention shall fall within the protection scope of the claims.

Claims (28)

An image data decoding method,
The image data decoding method is:
Obtaining a bitstream including image data;
Parsing the bitstream to obtain node split mode information of the first-level coding tree, wherein a root node of the first-level coding tree is assigned to one coding tree unit (CTU). And a leaf node of the first-level coding tree corresponds to a node split mode corresponding to the root node of the first-level coding tree and the node split mode information of the first-level coding tree. -Identified using;
Parsing the bitstream to obtain node division mode information of a second-level coding tree, wherein the node division mode information of the second-level coding tree corresponds to a first node of the second-level coding tree. Indicating a splitting mode, wherein the splitting mode corresponding to the first node is one mode in the candidate-split-mode set corresponding to the first node, and wherein the candidate-split-mode set corresponding to the first node is Determined according to a first preset partitioning condition, wherein the first preset partitioning condition is used to indicate whether the first node is to be restricted from splitting in a target partitioning mode, wherein the target partitioning mode is a horizontal binary partition. and at least one of a horizontal ternary split, a vertical binary split, and a vertical ternary split, wherein the root node of the second-level coding tree is the first-first. Leaf nodes of the tree, being Coding Bell;
If the splitting mode corresponding to the first node is no further splitting, parsing the bitstream to obtain encoding information of the first node, wherein the first node comprises one coding unit , Corresponding to CU)-; And
Decoding and reconstructing the coding unit based on the encoding information of the first node to obtain an image corresponding to the image data
Image data decoding method comprising a. The method of claim 1,
The node splitting mode corresponding to the first-level coding tree includes a quadtree split, and the node splitting mode corresponding to the second-level coding tree includes binary tree split and A method of decoding picture data comprising a ternary tree split. The method according to claim 1 or 2,
The first preset splitting condition is a preset splitting subcondition:
First preset division sub-condition: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to a first preset threshold, the candidate corresponding to the first node; The split-mode set does not include the horizontal binary division or the horizontal ternary division, wherein the width-to-height ratio of the image area corresponding to the first node is the image area corresponding to the first node. A condition that is the ratio of the width of to the height of the image area corresponding to the first node;
Second preset division sub-condition: if the height-to-width ratio of the picture region corresponding to the first node is greater than or equal to a second preset threshold, the candidate-division-mode set corresponding to the first node is the vertical; Not including binary division or the vertical ternary division, wherein the height to width ratio of the image region corresponding to the first node is equal to the height of the image region corresponding to the first node to the first node. A condition that is a ratio of the width of the corresponding image area;
Third preset division subcondition: a ratio of the area of the picture area corresponding to the first node to the area of the picture area corresponding to the leaf node of the first-level coding tree to which the first node belongs; If less than or equal to a preset threshold, the candidate-division-mode set corresponding to the first node does not include the horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. Condition;
Fourth preset partitioning subcondition: if the partitioning mode corresponding to the first node includes a first partitioning mode, decoding of the first child node of the first node is to decode the second child node of the first node. A later splitting mode corresponding to a second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode, where the first splitting mode is present. The splitting mode is the horizontal binary division or the vertical binary division;
Fifth preset division sub-condition: The division mode corresponding to the first node includes a second division mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the smallest of the immunities of the picture regions corresponding to three child nodes is small, the candidate-division-mode set corresponding to the first child node does not include the second partitioning mode, wherein the second partitioning mode is the A condition that is a horizontal ternary division or the vertical ternary division;
Sixth preset division subcondition: the division mode corresponding to the first node includes the second division mode, and the immunity of the image region corresponding to the first child node of the first node is determined by the first divisional mode. The largest of the immunities of the picture regions corresponding to the three child nodes of the node, the condition that the candidate-division-mode set corresponding to the first child node does not include the second division mode;
Seventh preset division subcondition: if the ratio of the width of the picture area corresponding to the first node to a preset side length of the minimum CU is less than or equal to a third preset threshold, The corresponding candidate-division-mode set does not include the vertical ternary division, or the ratio of the height of the picture region corresponding to the first node to the preset side length of the minimum CU is the third preset. If less than or equal to a threshold, the candidate-division-mode set corresponding to the first node comprises the horizontal ternary division;
Eighth preset partitioning subcondition: if the immunity of the picture region corresponding to the first node is less than or equal to a fourth preset threshold, the candidate-division-mode set corresponding to the first node is the horizontal binary. A condition not including division, the horizontal ternary division, the vertical binary division, or the vertical ternary division; And
9th preset partitioning subcondition: the partitioning mode corresponding to the first node is the second partitioning mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the immunity of the picture region corresponding to the three child nodes is the largest, the candidate-division-mode set corresponding to the first child node does not include the first partitioning mode, wherein The division direction is the same as the division direction of the second division
And at least one of the following. The method according to any one of claims 1 to 3,
The node division mode information of the second-level coding tree includes first information, second information, and third information, wherein the first information is used to indicate whether to further partition the first node, and the second Information is used to indicate a direction in which the first node is split, and the third information is used to indicate a mode in which the first node is split; And
Parsing the bitstream to obtain node division mode information of a second-level coding tree specifically:
Parsing the bitstream to determine, according to the first preset partitioning condition, the candidate-division-mode set corresponding to the first node; And
Parsing the bitstream based on a candidate-division-mode set corresponding to the first node to determine the first information, the second information, and the third information.
Image data decoding method comprising a. The method of claim 4, wherein
Parsing the bitstream based on the candidate-division-mode set corresponding to the first node to determine the first information, the second information, and the third information are specifically:
Determining a first numeric value, wherein the first numeric value is a quantity of partitioning modes included in the candidate-division-mode set corresponding to the first node; And
Determining the first information, the second information, and the third information by parsing the bitstream based on the first numeric value.
Image data decoding method comprising a. The method according to claim 4 or 5,
The node splitting mode corresponding to the second-level coding tree further includes the quadtree splitting; And
Parsing the bitstream to obtain node division mode information of a second-level coding tree specifically:
Parsing the bitstream to obtain the node split mode information of the second-level coding tree
Wherein the node splitting mode information of the second-level coding tree indicates the splitting mode corresponding to the first node of the second-level coding tree, and the mode corresponding to the first node The first node is one mode in the candidate-division-mode set corresponding to the first node, and the candidate-division-mode set corresponding to the first node is the first preset division condition and the second. Determined according to a preset dividing condition, wherein the second preset dividing condition is used to indicate whether the first node is to be restricted from splitting based on the quadtree splitting. The method of claim 6,
The second preset splitting condition is a preset splitting subcondition:
10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the candidate-division corresponding to the first node. A condition set does not include the quadtree splitting; And
Eleventh preset partitioning subcondition: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is divided into the quadtrees. Conditions that do not include
And at least one of the following. The method according to claim 6 or 7,
The node splitting mode information of the second-level coding tree further includes fourth information, the fourth information being used to indicate whether to split the first node based on the quadtree splitting; And
Parsing the bitstream to obtain node division mode information of a second-level coding tree specifically:
Parsing the bitstream to determine the candidate-division-mode set corresponding to the first node according to the first preset division condition and the second preset division condition; And
Parsing the bitstream based on the split mode corresponding to the first node to determine the first information, the second information, the third information, and the fourth information.
Image data decoding method comprising a. As the image data encoding method,
The picture encoding method is:
Determining a coding tree unit (CTU) corresponding to the picture block to be encoded;
Dividing the CTU in a node partitioning mode corresponding to a first-level coding tree to obtain a leaf node of the first-level coding tree-a root node of the first-level coding tree ) Corresponds to the CTU-;
Determining a candidate-division-mode set corresponding to the first node of the second-level coding tree, wherein the candidate-division-mode set corresponding to the first node satisfies a first preset partitioning condition; A first preset splitting condition is used to indicate whether the first node is to be restricted from splitting in a target splitting mode, the root node of the second-level coding tree is a leaf node of the first-level coding tree, and the target The splitting mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division; And
If the candidate-division-mode set corresponding to the first node includes non-additional division, a coding unit bitstream corresponding to the coding unit by encoding a coding unit (CU) corresponding to the first node. Obtaining
Image data encoding method comprising a. The method of claim 9,
The node splitting mode corresponding to the first-level coding tree includes quadtree splitting, and the node splitting mode corresponding to the second-level coding tree includes binary tree splitting and ternary tree splitting. Data encoding method. The method according to claim 9 or 10,
The first preset splitting condition is a preset splitting subcondition:
First preset division sub-condition: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to a first preset threshold, the candidate corresponding to the first node; The split-mode set does not include the horizontal binary division or the horizontal ternary division, wherein the width-to-height ratio of the image area corresponding to the first node is the image area corresponding to the first node. A condition that is the ratio of the width of to the height of the image area corresponding to the first node;
Second preset division sub-condition: if the height-to-width ratio of the picture region corresponding to the first node is greater than or equal to a second preset threshold, the candidate-division-mode set corresponding to the first node is the vertical; Not including binary division or the vertical ternary division, wherein the height to width ratio of the image region corresponding to the first node is equal to the height of the image region corresponding to the first node to the first node. A condition that is a ratio of the width of the corresponding image area;
Third preset division subcondition: a ratio of the area of the picture area corresponding to the first node to the area of the picture area corresponding to the leaf node of the first-level coding tree to which the first node belongs; If less than or equal to a preset threshold, the candidate-division-mode set corresponding to the first node does not include the horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. Condition;
Fourth preset partitioning subcondition: if the partitioning mode corresponding to the first node includes a first partitioning mode, decoding of the first child node of the first node is to decode the second child node of the first node. A later splitting mode corresponding to a second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode, where the first splitting mode is present. The splitting mode is the horizontal binary division or the vertical binary division;
Fifth preset division sub-condition: The division mode corresponding to the first node includes a second division mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the smallest of the immunities of the picture regions corresponding to three child nodes is small, the candidate-division-mode set corresponding to the first child node does not include the second partitioning mode, wherein the second partitioning mode is the A condition that is a horizontal ternary division or the vertical ternary division;
Sixth preset division subcondition: the division mode corresponding to the first node includes the second division mode, and the immunity of the image region corresponding to the first child node of the first node is determined by the first divisional mode. The largest of the immunities of the picture regions corresponding to the three child nodes of the node, the condition that the candidate-division-mode set corresponding to the first child node does not include the second division mode;
Seventh preset division subcondition: if the ratio of the width of the picture area corresponding to the first node to a preset side length of the minimum CU is less than or equal to a third preset threshold, The corresponding candidate-division-mode set does not include the vertical ternary division, or the ratio of the height of the picture region corresponding to the first node to the preset side length of the minimum CU is the third preset. If less than or equal to a threshold, the candidate-division-mode set corresponding to the first node comprises the horizontal ternary division;
Eighth preset partitioning subcondition: if the immunity of the picture region corresponding to the first node is less than or equal to a fourth preset threshold, the candidate-division-mode set corresponding to the first node is the horizontal binary. A condition not including division, the horizontal ternary division, the vertical binary division, or the vertical ternary division; And
9th preset partitioning subcondition: the partitioning mode corresponding to the first node is the second partitioning mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the immunity of the picture region corresponding to the three child nodes is the largest, the candidate-division-mode set corresponding to the first child node does not include the first partitioning mode, wherein the first partitioning mode The division direction is the same as the division direction of the second division
And at least one of the following. The method of claim 10,
The node splitting mode corresponding to the second-level coding tree further includes the quadtree splitting; And
Determining the candidate-division-mode set corresponding to the first node of the second-level coding tree,
Determining the candidate-division-mode set corresponding to the first node
More,
The candidate-division-mode set corresponding to the first node satisfies the first preset partitioning condition and the second preset partitioning condition, wherein the second preset partitioning condition is determined by the first node in the quadtree. Used to indicate whether to restrict the division based on the division. The method of claim 12,
The second preset splitting condition is a preset splitting subcondition:
10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the candidate-division corresponding to the first node. A condition set does not include the quadtree splitting; And
Eleventh preset partitioning subcondition: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is divided into the quadtrees. Conditions that do not include
And at least one of the following. The method according to any one of claims 9 to 13,
The image data encoding method,
If the candidate-division-mode set corresponding to the first node includes a split mode other than no further splitting, then the rate distortion cost of each split mode available to the first node Calculating;
Determining a splitting mode corresponding to a minimum rate distortion cost as a target splitting mode corresponding to the first node; And
Dividing the first node in a target splitting mode corresponding to the first node
Image data encoding method further comprising. As a decoder device,
An acquisition module, configured to acquire a bitstream including the image data;
Parse the bitstream to obtain node split mode information of a first-level coding tree, wherein a root node of the first-level coding tree corresponds to one coding tree unit (CTU) And a leaf node of the first-level coding tree uses a node split mode corresponding to the root node of the first-level coding tree and the node split mode information of the first-level coding tree. Identified by-; Parse the bitstream to obtain node split mode information of a second-level coding tree, wherein the node split mode information of the second-level coding tree corresponds to a first node of the second-level coding tree Indicating a mode, wherein the splitting mode corresponding to the first node is one mode in the candidate-dividing-mode set corresponding to the first node, and wherein the candidate-dividing-mode set corresponding to the first node is the first mode. 1 is determined according to a preset division condition, and the first preset division condition is used to indicate whether the first node is to be restricted from splitting in a target division mode, wherein the target division mode is a horizontal binary split. ), At least one of a horizontal ternary split, a vertical binary split or a vertical ternary split, wherein the root node of the second-level coding tree is the first-level. Leaf node of the tree being coded; And a parsing module, configured to parse the bitstream to obtain encoding information of the first node if the splitting mode corresponding to the first node is no further splitting. Corresponds to a coding unit (CU); And
A decoding and reconstruction module, configured to decode and reconstruct the coding unit based on the encoding information of the first node to obtain a picture corresponding to the picture data
Decoder device comprising a. The method of claim 15,
Wherein the node splitting mode corresponding to the first-level coding tree comprises quadtree splitting and the node splitting mode corresponding to the second-level coding tree comprises binary tree splitting and ternary tree splitting. Device. The method according to claim 15 or 16,
The first preset splitting condition is a preset splitting subcondition:
First preset division sub-condition: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to a first preset threshold, the candidate corresponding to the first node; The split-mode set does not include the horizontal binary division or the horizontal ternary division, wherein the width-to-height ratio of the image area corresponding to the first node is the image area corresponding to the first node. A condition that is the ratio of the width of to the height of the image area corresponding to the first node;
Second preset division sub-condition: if the height-to-width ratio of the picture region corresponding to the first node is greater than or equal to a second preset threshold, the candidate-division-mode set corresponding to the first node is the vertical; Not including binary division or the vertical ternary division, wherein the height to width ratio of the image region corresponding to the first node is equal to the height of the image region corresponding to the first node to the first node. A condition that is a ratio of the width of the corresponding image area;
Third preset division subcondition: a ratio of the area of the picture area corresponding to the first node to the area of the picture area corresponding to the leaf node of the first-level coding tree to which the first node belongs; If less than or equal to a preset threshold, the candidate-division-mode set corresponding to the first node does not include the horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. Condition;
Fourth preset partitioning subcondition: if the partitioning mode corresponding to the first node includes a first partitioning mode, decoding of the first child node of the first node is to decode the second child node of the first node. A later splitting mode corresponding to a second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode, where the first splitting mode is present. The splitting mode is the horizontal binary division or the vertical binary division;
Fifth preset division sub-condition: The division mode corresponding to the first node includes a second division mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the smallest of the immunities of the picture regions corresponding to three child nodes is small, the candidate-division-mode set corresponding to the first child node does not include the second partitioning mode, wherein the second partitioning mode is the A condition that is a horizontal ternary division or the vertical ternary division;
Sixth preset division subcondition: the division mode corresponding to the first node includes the second division mode, and the immunity of the image region corresponding to the first child node of the first node is determined by the first divisional mode. The largest of the immunities of the picture regions corresponding to the three child nodes of the node, the condition that the candidate-division-mode set corresponding to the first child node does not include the second division mode;
Seventh preset division subcondition: if the ratio of the width of the picture area corresponding to the first node to a preset side length of the minimum CU is less than or equal to a third preset threshold, The corresponding candidate-division-mode set does not include the vertical ternary division, or the ratio of the height of the picture region corresponding to the first node to the preset side length of the minimum CU is the third preset. If less than or equal to a threshold, the candidate-division-mode set corresponding to the first node comprises the horizontal ternary division;
Eighth preset partitioning subcondition: if the immunity of the picture region corresponding to the first node is less than or equal to a fourth preset threshold, the candidate-division-mode set corresponding to the first node is the horizontal binary. A condition not including division, the horizontal ternary division, the vertical binary division, or the vertical ternary division; And
9th preset partitioning subcondition: the partitioning mode corresponding to the first node is the second partitioning mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the immunity of the picture region corresponding to the three child nodes is the largest, the candidate-division-mode set corresponding to the first child node does not include the first partitioning mode, wherein the first partitioning mode The division direction is the same as the division direction of the second division
A decoder device comprising at least one of. The method according to any one of claims 15 to 17,
The node division mode information of the second-level coding tree includes first information, second information, and third information, wherein the first information is used to indicate whether to further partition the first node, and the second Information is used to indicate a direction in which the first node is split, and the third information is used to indicate a mode in which the first node is split; And
The parsing module specifically: parses the bitstream to determine, according to the first preset partitioning condition, the candidate-division-mode set corresponding to the first node; And parse the bitstream based on a candidate-division-mode set corresponding to the first node to determine the first information, the second information, and the third information. The method of claim 18,
The parsing module specifically determines: a first numeric value, the first numeric value being a quantity of partitioning modes included in the candidate-division-mode set corresponding to the first node; And parse the bitstream based on the first numeric value to determine the first information, the second information, and the third information. The method of claim 18 or 19,
The node splitting mode corresponding to the second-level coding tree further includes the quadtree splitting; And
The parsing module is specifically configured to: parse the bitstream to obtain the node split mode information of the second-level coding tree, wherein the node split mode information of the second-level coding tree Indicate the splitting mode corresponding to the first node of a second-level coding tree, wherein the mode corresponding to the first node is one in the candidate-dividing-mode set corresponding to the first node; The candidate-division-mode set corresponding to the first node is determined according to the first preset division condition and the second preset division condition, and the second preset division condition is the first node. Is used to indicate whether to restrict splitting based on the quadtree splitting. The method of claim 20,
The second preset splitting condition is a preset splitting subcondition:
10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the candidate-division corresponding to the first node. A condition set does not include the quadtree splitting; And
Eleventh preset partitioning subcondition: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is divided into the quadtrees. Conditions that do not include
A decoder device comprising at least one of. The method of claim 20 or 21,
The node splitting mode information of the second-level coding tree further includes fourth information, the fourth information being used to indicate whether to split the first node based on the quadtree splitting; And
The parsing module specifically: parses the bitstream to determine the candidate-division-mode set corresponding to the first node according to the first preset division condition and the second preset division condition; And parse the bitstream based on a split mode corresponding to the first node to determine the first information, the second information, the third information, and the fourth information. As an encoder device,
A determining module, configured to determine a coding tree unit (CTU) corresponding to the picture block to be encoded;
A splitting module, configured to split the CTU in a node splitting mode corresponding to a first-level coding tree to obtain a leaf node of the first-level coding tree-a root of the first-level coding tree A node (root node) corresponds to the CTU, and the determination module is further configured to determine a candidate-division-mode set corresponding to the first node of the second-level coding tree and corresponding to the first node. Wherein the candidate-division-mode set satisfies a first preset dividing condition, wherein the first preset dividing condition is used to indicate whether the first node is to be restricted from dividing in a target dividing mode, and wherein the second-level The root node of the coding tree is a leaf node of the first-level coding tree, and the target division mode includes at least one of horizontal binary division, horizontal ternary division, vertical binary division, and vertical ternary division. Also; And
A coding unit corresponding to the first node if the candidate-division-mode set corresponding to the first node and the candidate-division-mode set determined by the determining module includes non-additional division. , An encoding module, configured to encode a CU) to obtain a coding unit bitstream corresponding to the coding unit
Encoder device comprising a. The method of claim 23, wherein
The node splitting mode corresponding to the first-level coding tree comprises quadtree splitting, and the node splitting mode corresponding to the second-level coding tree comprises binary tree splitting and ternary tree splitting. Device. The method of claim 23 or 24,
The first preset splitting condition is a preset splitting subcondition:
First preset division sub-condition: if the width-to-height ratio of the picture area corresponding to the first node is greater than or equal to a first preset threshold, the candidate corresponding to the first node; The split-mode set does not include the horizontal binary division or the horizontal ternary division, wherein the width-to-height ratio of the image area corresponding to the first node is the image area corresponding to the first node. A condition that is the ratio of the width of to the height of the image area corresponding to the first node;
Second preset division sub-condition: if the height-to-width ratio of the picture region corresponding to the first node is greater than or equal to a second preset threshold, the candidate-division-mode set corresponding to the first node is the vertical; Not including binary division or the vertical ternary division, wherein the height to width ratio of the image region corresponding to the first node is equal to the height of the image region corresponding to the first node to the first node. A condition that is a ratio of the width of the corresponding image area;
Third preset division subcondition: a ratio of the area of the picture area corresponding to the first node to the area of the picture area corresponding to the leaf node of the first-level coding tree to which the first node belongs; If less than or equal to a preset threshold, the candidate-division-mode set corresponding to the first node does not include the horizontal binary division, the horizontal ternary division, the vertical binary division, or the vertical ternary division. Condition;
Fourth preset partitioning subcondition: if the partitioning mode corresponding to the first node includes a first partitioning mode, decoding of the first child node of the first node is to decode the second child node of the first node. A later splitting mode corresponding to a second child node is the first splitting mode, and the candidate-split-mode set corresponding to the first child node does not include the first splitting mode, where the first splitting mode is present. The splitting mode is the horizontal binary division or the vertical binary division;
Fifth preset division sub-condition: The division mode corresponding to the first node includes a second division mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the smallest of the immunities of the picture regions corresponding to three child nodes is the smallest, the candidate-division-mode set corresponding to the first child node does not include the second division mode, wherein the second division mode is A condition that is a horizontal ternary division or the vertical ternary division;
Sixth preset division subcondition: the division mode corresponding to the first node includes the second division mode, and the immunity of the image region corresponding to the first child node of the first node is determined by the first divisional mode. The largest of the immunities of the picture regions corresponding to the three child nodes of the node, the condition that the candidate-division-mode set corresponding to the first child node does not include the second division mode;
Seventh preset division subcondition: if the ratio of the width of the picture area corresponding to the first node to a preset side length of the minimum CU is less than or equal to a third preset threshold, The corresponding candidate-division-mode set does not include the vertical ternary division, or the ratio of the height of the picture region corresponding to the first node to the preset side length of the minimum CU is the third preset. If less than or equal to a threshold, the candidate-division-mode set corresponding to the first node comprises the horizontal ternary division;
Eighth preset partitioning subcondition: if the immunity of the picture region corresponding to the first node is less than or equal to a fourth preset threshold, the candidate-division-mode set corresponding to the first node is the horizontal binary. A condition not including division, the horizontal ternary division, the vertical binary division, or the vertical ternary division; And
9th preset partitioning subcondition: the partitioning mode corresponding to the first node is the second partitioning mode, and the immunity of the image area corresponding to the first child node of the first node is equal to that of the first node. If the immunity of the picture region corresponding to the three child nodes is the largest, the candidate-division-mode set corresponding to the first child node does not include the first partitioning mode, wherein the first partitioning mode The division direction is the same as the division direction of the second division
An encoder device comprising at least one of. The method of claim 24,
The node splitting mode corresponding to the second-level coding tree further includes the quadtree splitting; And
The determining module is specifically configured to determine the candidate-division-mode set corresponding to the first node, wherein the candidate-division-mode set corresponding to the first node is the first preset partitioning condition. And satisfying the second preset splitting condition, wherein the second preset splitting condition is used to indicate whether the first node is to be restricted from splitting based on the quadtree splitting. The method of claim 26,
The second preset splitting condition is a preset splitting subcondition:
10th preset partitioning subcondition: if the depth of the first node in the second-level coding tree is less than a preset maximum depth in the second-level coding tree, the candidate-division corresponding to the first node. A condition set does not include the quadtree splitting; And
Eleventh preset partitioning subcondition: if the width-to-height ratio of the picture area corresponding to the first node is greater than a fifth preset threshold, the candidate-division-mode set corresponding to the first node is divided into the quadtrees. Conditions that do not include
An encoder device comprising at least one of. The method according to any one of claims 23 to 27,
The encoder device further includes a calculation module,
The calculating module comprises: the candidate-split-mode set corresponding to the first node and wherein the candidate-split-mode set determined by the determining module includes a split mode that is not no further splitting. And calculate the rate distortion cost of each splitting mode available to the first node,
The determining module is further configured to determine a splitting mode corresponding to the minimum rate distortion cost as a target splitting mode corresponding to the first node, and
The splitting module is specifically configured to split the first node in a target splitting mode determined by the determining module while being a target splitting mode corresponding to the first node.

Images